What Is Restaking? EigenLayer, LRTs, and the 2026 Reset

Restaking was the loudest idea in Ethereum DeFi through 2024 and 2025. The pitch sounded like a free upgrade: the same staked ETH that already secures Ethereum could be pledged a second time to secure other services, and earn a second yield for the trouble. Two years and one nine-figure exploit later, the picture is a good deal more sober.

This is a plain-English guide to what restaking actually is, how the yield stack is assembled, who the main players are in mid-2026, what broke in April 2026, and where the SEC has drawn its lines. Prices are in USD, and none of this is financial advice.

Restaking in one sentence

Restaking lets you re-pledge ETH that is already staked (or a token that represents staked ETH) so it also backs third-party services such as oracles, bridges, data-availability layers, and rollup sequencers. Those services, called Actively Validated Services (AVS) in EigenLayer’s vocabulary, borrow economic security from Ethereum’s validator set instead of bootstrapping their own from scratch. In exchange, restakers collect extra rewards, and they accept extra slashing conditions: if a restaked validator misbehaves in a service it opted into, its stake can be cut.

The word was coined by EigenLayer, which built the first restaking marketplace on Ethereum. By mid-2026 it describes a whole category, complete with rivals, liquid wrappers, and a systemic-risk track record of its own.

From staking to restaking: how the yield stack is built

Picture Ethereum yield as a stack of layers, each adding return and risk on top of the one below. The base layer is ordinary staking: lock 32 ETH (or delegate to a pool), run a validator, and earn a native rate of roughly 2.78% plus a little from MEV. Around 32% of all ETH is staked across some 897,000 validators, so this is a deep, well-understood market.

Liquid staking tokens (LSTs) like stETH sit one layer up: they hand you a tradable receipt so your capital is not frozen while it stakes. Restaking adds the next layer: you deposit ETH or an LST into a restaking protocol, delegate it to a professional operator, and that operator runs software for one or more AVS that pay their own rewards. Liquid restaking tokens (LRTs) then wrap the restaked position into yet another liquid token you can redeploy elsewhere in DeFi. Every layer stacks yield, and every layer stacks a fresh failure mode.

EigenLayer, the protocol that coined the term

EigenLayer rebranded to EigenCloud in 2025 as it widened its pitch from pure restaking toward a broader verifiable-cloud story of data availability, coprocessors, and offchain compute. The pivotal moment came on 17 April 2025, when slashing went live on mainnet on an opt-in basis. That is when restaking stopped being a promise about future penalties and became a live one. By mid-2026 the protocol counts roughly 190 AVS in development, about 40 on mainnet, more than 2,000 operators, and over 80,000 addresses.

The money tells the harder half of the story. Total value locked peaked near $19.7 billion during the 2024 points-farming frenzy, fell to about $8.9 billion by March 2026, and kept sliding to roughly $4.7 billion by mid-2026, according to DefiLlama. The EIGEN token trades near $0.21 with a market capitalization around $155 million and a circulating supply close to 740 million; its all-time high was $5.65 in December 2024, leaving it down about 96%, per CoinGecko.

Liquid restaking tokens and rehypothecation

Restaking locks your ETH behind the services it secures, so LRTs exist to give that capital its liquidity back. Deposit into a liquid restaking protocol and it mints you a token (eETH, ezETH, rsETH, pufETH) that represents your restaked position. You can trade it, lend it, or post it as collateral while it keeps earning. That reuse, known as rehypothecation, is the source of both the extra yield and the extra fragility: the same token can be doing three jobs at once across three protocols.

ether.fi (eETH) is the clear leader with more than $3.2 billion, followed by Renzo (ezETH) near $2 billion, Puffer (pufETH) around $1.3 billion, and Kelp DAO (rsETH) above $740 million. The segment first vaulted toward $8 billion in total value locked during its 2024 growth spurt. Yields have historically run from the high single digits into the low teens in percent, but as DL News flagged early on, those rewards ride on layered smart-contract and depeg risk; ezETH briefly lost its peg in April 2024 and liquidated leveraged positions.

The Kelp DAO hack and how the risk spread to Aave

The clearest illustration of restaking’s systemic edge arrived on Saturday, 19 April 2026. An attacker exploited Kelp DAO’s cross-chain bridge to mint 116,500 rsETH out of thin air, worth about $292 million and equal to roughly 18% of the token’s circulating supply. It stands as the largest DeFi exploit of 2026, and the damage did not stay contained to Kelp.

Because rsETH was accepted as collateral elsewhere, the attacker deposited the freshly minted tokens on Aave V3 and borrowed wrapped ETH against them, leaving the lending market with roughly $196 million in bad debt. Aave’s total value locked fell from $26.4 billion to about $20 billion over a single weekend, and the AAVE token dropped around 16% to $92, as CoinDesk reported. The lesson landed hard: a money market had taken an LRT as collateral without fully pricing the bridge and redemption risk stacked underneath it. Restaking risk is composable, and so is restaking contagion.

Vitalik’s warning: do not overload Ethereum’s consensus

None of this caught Ethereum’s designers by surprise. Two years before the Kelp hack, co-founder Vitalik Buterin published “Don’t overload Ethereum’s consensus” on 21 May 2023, in effect a preemptive critique of restaking. His worry was that pushing validators to adjudicate more and more external questions lets disputes in those side services leak back into Ethereum’s core consensus and split the community, while the sheer size of a restaking failure creates pressure to bail it out rather than let it burn.

EigenLayer co-founder Sreeram Kannan agreed on the narrow point of not loading consensus with complex financial primitives, and the architecture tries to keep each AVS failure walled off from the base chain. Whether those walls hold under real stress is precisely the question April 2026 put to the test.

The competition: Symbiotic, Karak, and SSV’s based applications

EigenLayer no longer has the field to itself. Symbiotic went live on mainnet on 28 January 2025 with slashing from day one; it accepts any ERC-20 as collateral rather than only ETH, and it farms dispute resolution out to external resolvers such as UMA, Kleros, or a chosen committee. Backed by a $29 million round led by Pantera, it has attracted around $1.7 billion. Karak rounds out the general-purpose field as a third multi-asset venue.

SSV Network, Ethereum’s largest distributed-validator provider, is pushing a deliberately different design it calls based applications. Rather than restaking the 32 ETH, the existing validator set lends its security and liveness to applications while that principal stays untouched and unslashable; only optional token delegations are ever at risk. SSV frames this as an infinite-sum model against restaking’s zero-sum one, a direct jab at the idea that the same collateral should keep absorbing more and more slashing conditions.

Where the SEC stands on staking and restaking

Under Chair Paul Atkins, the SEC has eased its posture on staking. On 29 May 2025 the Division of Corporation Finance issued a staff statement concluding that certain protocol staking activities (solo, delegated, and custodial staking of assets used to run a network) are not securities transactions, because the work is administrative or ministerial rather than the entrepreneurial efforts of others that the Howey test targets. Commissioner Hester Peirce backed it with a companion note titled “Providing Security is not a Security.”

On 5 August 2025 the staff went further, extending the shelter to liquid staking and staking receipt tokens of the stETH and eETH variety in a follow-up Peirce dubbed the staking sequel; the reasoning is that such a token merely evidences the deposited asset and is not itself a security. Commissioner Caroline Crenshaw dissented from both statements.

Here is the caveat that matters most for restaking: none of that guidance clearly reaches restaking or the more complex liquid restaking tokens. Once the yield depends on an operator actively running a business (picking AVS, managing strategies, and taking a cut), the return starts to resemble a share of someone else’s enterprise, which is exactly what Howey scrutinizes. All of it is staff-level and non-binding, and it predates any restaking test case. The bigger market-structure question sits with Congress: the CLARITY Act passed the House on 17 July 2025 and cleared the Senate Banking Committee by 15 votes to 9 on 14 May 2026, and it would divide spot-crypto oversight between the SEC and the CFTC, though 2026 passage is far from certain.

Should you restake? What to weigh

Restaking is a genuine primitive with genuine uses, but in 2026 the market prices it as the higher-risk end of the yield stack, not the free lunch that 2024 marketing implied. Before committing capital, weigh a few things:

• Slashing has been live and real since April 2025, not a theoretical future penalty.
• LRTs layer smart-contract, bridge, and depeg risk on top of ordinary staking risk.
• Rehypothecation means one protocol’s failure can cascade into lending markets, as Kelp and Aave showed.
• Yields have compressed as TVL and AVS fee revenue fell short of the points-era hype.
• US regulatory comfort covers plain staking, not restaking or complex LRTs.

Read each protocol’s audits and redemption mechanics, size positions to what you can afford to lose, and remember that a liquid wrapper is only ever as solid as the market’s willingness to price it at par. None of this is financial advice.

By the HOGE Wire editorial desk.