Restaking Explained: EigenLayer, LRTs and Shared Security

Restaking is one of the biggest ideas to come out of Ethereum since liquid staking, and one of the most argued over. In less than three years it pulled tens of billions of dollars in staked ETH into a new layer of the DeFi stack, produced a family of tokens with names like weETH and ezETH, drew a public warning from Vitalik Buterin, and, in April 2026, sat at the center of the year’s single largest crypto exploit. This explainer covers what restaking is, how it works, what it pays, and why serious people still disagree about whether it is a breakthrough or a hazard.

What restaking actually means

Start with plain staking. To help secure Ethereum, holders lock ETH with a validator and earn a yield for honest work, currently around 3 percent a year. That ETH sits idle apart from its one job: keeping the network safe. Restaking asks a simple question. If the capital is already bonded and can be slashed for misbehavior, why not let it secure a second service at the same time, for extra pay?

That is the whole idea. Restaking takes ETH that is already staked and pledges it again to back other protocols, such as bridges, oracles, and data availability layers, that need their own economic security. The projects that pioneered the model call those customers Actively Validated Services, or AVSs. The dominant platform, EigenLayer (rebranded EigenCloud in 2025), turned the concept into a multibillion dollar market in a matter of months.

Giving your ETH a second job

EigenLayer grew out of research by Sreeram Kannan and his team at Eigen Labs, with a whitepaper in 2023. The pitch to founders was blunt: Ethereum already has an enormous pool of staked ETH securing it, so why should every new middleware project bootstrap its own token and validator set from scratch, an approach that is expensive and, early on, weakly secured? Restaking lets those projects rent Ethereum’s existing security instead.

In exchange, AVSs pay rewards to two groups: the restakers who supply the capital and the operators who run the software. It is a marketplace. Restakers bring bonded ETH, operators bring hardware and uptime, and each AVS brings a budget plus a set of rules for what counts as misbehavior. Get it right and a new protocol inherits billions of dollars of security on day one instead of spending years building it.

Native restaking versus liquid restaking

There are two front doors. Native restaking is for people who run their own Ethereum validators; they point their withdrawal credentials at EigenLayer smart contracts, keeping the ETH staked on Ethereum while opting in to secure AVSs. Liquid restaking is for everyone else: you deposit a liquid staking token such as Lido’s stETH or Rocket Pool’s rETH, and the protocol restakes it on your behalf.

Either way, most restakers delegate to a professional operator rather than run AVS software themselves. The piece that was missing for a long time was real punishment. Slashing, the mechanism that lets an AVS confiscate a misbehaving operator’s stake, only went live on EigenLayer’s mainnet on April 17, 2025. That is the moment the security stopped being theoretical and started carrying teeth.

Liquid restaking tokens, the retail on-ramp

Most of the money did not arrive through validators. It came through liquid restaking tokens, or LRTs. A protocol takes your ETH or staking token, restakes it, and issues a new liquid token that represents the position. You can trade that token, lend it, or post it as collateral while it keeps earning, which is why LRTs spread quickly across DeFi. ether.fi’s weETH became the largest, followed by Renzo’s ezETH and Kelp DAO’s rsETH.

Those figures are approximate, move with the price of ETH, and come from DefiLlama; treat them as a snapshot rather than a live quote.

The yield, and where it really comes from

The appeal is stacked yield. Through a good stretch of 2024 and 2025, an LRT could advertise a base staking return of roughly 3 percent, plus AVS rewards, plus points that protocols dangled ahead of token airdrops, for headline numbers in the 8 to 12 percent range. Some aggressive loop strategies claimed a good deal more.

Read those numbers with care. A large share of the early boom was points farming, deposits chasing speculative airdrops rather than real cash flow. AVSs are still young, and the fees they actually pay for security remain modest, so the durable, cash-backed portion of restaking yield is much smaller than the flashiest APYs suggest. When the points music slows, so does most of the return.

The risks you cannot restake away

Restaking does not conjure yield from nothing; it sells extra risk, and that risk arrives in layers rather than as a single number on a dashboard.

• Slashing risk: one pool of ETH backs many services, so penalties can stack across them.
• Smart contract risk: the AVS, the restaking layer, and the LRT wrapper each add attack surface.
• Bridge and infrastructure risk: wrapped LRTs lean on cross-chain messaging that can fail on its own.
• Liquidity and depeg risk: an LRT can trade below the value of the assets behind it when holders rush the exit.
• Systemic risk: heavy reliance on one platform concentrates failure, the core of Buterin’s warning.

Vitalik Buterin flagged that deeper worry back in a 2023 essay, Don’t overload Ethereum’s consensus. He warned that piling extra duties onto staked ETH could import systemic risk into the base layer and breed too-big-to-fail dynamics, where the failure of a large restaking service creates pressure to bail it out through Ethereum itself. His advice was to keep the base chain minimal and treat restaking with caution.

The Kelp DAO hack and the bridge problem

That caution got a hard illustration in the spring of 2026. On April 19, an attacker drained about 116,500 rsETH, worth roughly 292 million dollars, from a Kelp DAO bridge, then watched wrapped versions of the token strand across more than 20 chains as their backing vanished. It was the largest DeFi exploit of the year, and it froze rsETH lending markets on Aave, Spark, and Fluid within hours.

Here is the detail that matters: this was not a break in restaking math or in slashing. According to Chainalysis, attackers linked to North Korea’s Lazarus Group subverted the protocol’s cross-chain messaging setup, a single-point-of-failure verification path, to release tokens that were never properly backed. Restaking stacks contracts, bridges, and off-chain infrastructure on top of one another, and Kelp showed that the weakest link, in this case a bridge, can drain the whole position no matter how sound the core design is.

Restaking beyond Ethereum

EigenLayer no longer has the field to itself. Symbiotic, backed by the venture firm Paradigm, takes a more permissionless, multi-asset approach and had drawn several hundred million dollars in deposits by mid-2026, while Karak spreads restaking across multiple chains and collateral types. Competition has pushed all of them to court AVSs with better terms and clearer slashing rules.

The idea has also jumped ecosystems. On Solana, Jito built restaking infrastructure around JitoSOL and what it calls Node Consensus Networks, letting staked SOL and other tokens secure extra services much as EigenLayer does on Ethereum. Shared security, it turns out, is not an Ethereum-only story.

Where US regulators stand

For a US audience, the regulatory backdrop matters as much as the yield. The Securities and Exchange Commission spent 2023 treating staking services with suspicion. It fined Kraken 30 million dollars that February and forced it to shut its US staking-as-a-service program, calling it an unregistered securities offering.

The tone changed in 2025. The SEC’s Division of Corporation Finance issued a statement on May 29, 2025 that plain protocol staking is not a securities transaction, and followed it on August 5 with similar comfort for certain liquid staking activities. Two cautions apply. These are staff statements, not laws or formal rules, and they address staking, not the extra machinery of restaking: delegation to operators, AVS reward flows, and LRT wrappers. Plain staking has clearer footing in the United States today than restaking does.

What to watch next

EigenLayer’s own answer to the question of what all this is for has been to broaden out. In 2025 it rebranded as EigenCloud and, with a fresh 70 million dollar investment from a16z, repositioned as a verifiable cloud spanning data availability (EigenDA), off-chain compute, and dispute resolution, with verifiable AI as a headline use case.

The market, meanwhile, has cooled. Restaking value locked slid from a peak above 20 billion dollars to the mid-single-digit and low-double-digit billions by mid-2026, and the EIGEN token traded near 21 cents in July, far below its debut. The open question for the rest of 2026 is whether AVSs can generate enough real fee revenue to pay for the security they consume, which would make restaking yield durable, or whether it stays a narrative that ran ahead of its cash flows. Watch three things: AVS fees, the size and frequency of slashing and exploit events, and whether US regulators extend their new staking comfort to restaking’s extra layers.

By the HOGE Wire desk. This is market commentary, not investment advice; restaking carries a real risk of losing principal, so do your own research before committing any assets.