DeFi Intel
Home · Guides · Liquid Restaking Tokens (LRTs) Explained: ether.fi, Renzo, Kelp, Swell, Puffer, Mellow, and What Survives 2026

Liquid Restaking Tokens (LRTs) Explained: ether.fi, Renzo, Kelp, Swell, Puffer, Mellow, and What Survives 2026

Liquid restaking tokens (LRTs) emerged in 2023 as a derivative layer on top of EigenLayer restaking (Eigen Labs rebranded its broader platform to EigenCloud in mid-2025; the restaking layer LRTs build on is still commonly called EigenLayer) — a way to deposit ETH or LSTs into restaking, capture all the resulting yields and points, and receive a tradeable token usable as collateral or composable across DeFi. The narrative caught fire through 2024, with combined LRT TVL crossing $13B at peak. Following the EIGEN airdrop and the wind-down of the points era, the market has consolidated to a handful of credible issuers managing $7-9B in steady-state TVL. This guide walks through what LRTs are, how each major issuer is differentiated, the points-era economics that drove the boom, the post-points yield compression, the structural risks, and which players survive 2026's continued consolidation.

What an LRT actually is

The deposit-and-mint mechanic

A user deposits ETH (or a liquid staking token like stETH or wstETH) into an LRT issuer's contract. The contract performs three operations: (1) stakes the ETH (if not already an LST) through its preferred validator infrastructure, (2) restakes the resulting LST or ETH-validator into EigenLayer, delegating to a curated set of AVSs, and (3) mints LRT tokens to the depositor representing their proportional share of the restaked position.

The LRT accrues value as the underlying restaked position generates yield — base ETH staking + AVS rewards + any LRT-issuer points. The LRT is freely transferable, usable as collateral on lending protocols (Aave, Morpho), tradeable on DEXs, and composable across DeFi (Pendle PT/YT markets, Curve pools, Balancer weighted pools).

Why a separate token vs direct EigenLayer position

Direct EigenLayer restaking requires running validators (operationally complex), choosing AVS delegations (expertise-heavy), and managing withdrawal queues (illiquid). An LRT abstracts this: the issuer runs the operator infrastructure and selects AVSs; the user holds a tradeable claim. This is identical to the value liquid staking tokens (Lido stETH) provide over running a solo validator — operational simplicity and DeFi composability in exchange for fees and operator trust.

The major LRT issuers

ether.fi (eETH / weETH)

Largest LRT by TVL through 2024-2026 — typically $3-4B. ether.fi distinguishes itself with a delegated-validator model: stakers can opt to run their own validator infrastructure under ether.fi's stack rather than fully entrusting operations to ether.fi-controlled operators. This gives the protocol a solo-staker-friendly story and a more decentralized operator profile than Lido.

ETHFI token launched in 2024 with airdrops to LRT participants; trades as governance/value-capture token for the protocol. weETH is the wrapped, non-rebasing version of eETH used for DeFi composability. Pendle eETH markets were the deepest LRT yield-stripping market during the points era.

Renzo (ezETH)

Second-largest LRT by TVL, peaked in early 2024 with aggressive points distribution and multi-chain expansion (Ethereum, Arbitrum, Linea, Mode, Blast). Renzo's strategy emphasized integration breadth and points marketing.

In April 2024, ezETH suffered a depeg event triggered by a token unlock and concentrated exit on a Curve pool — the price fell sharply below peg on secondary markets before recovering, punishing leveraged ezETH borrowers and triggering several lending-protocol liquidations. The post-mortem became a case study in LRT liquidity risk: AMM pools backing LRT trading need depth proportional to leveraged collateral usage. Renzo retained its top-tier position post-depeg but the event tempered the perception of LRT peg stability.

Kelp DAO (rsETH)

Kelp accepts a wider range of deposits — stETH, ETHx (Stader), sfrxETH (Frax) — rather than only ETH, giving it a broader collateral acceptance profile. This means rsETH effectively wraps existing LST positions into restaking, useful for users who already hold an LST and don't want to unwind. KEP token launched in 2024-2025 with allocation to early depositors. rsETH TVL typically sits at $1-2B. In April 2026 Kelp suffered the largest DeFi exploit of the year: roughly $292M was drained from its LayerZero-powered cross-chain bridge on 19 April 2026 (about 116,500 rsETH, ~18 percent of supply), traced to a 1-of-1 message-verifier configuration and attributed to the Lazarus Group. The stolen rsETH was used as collateral on Aave, which froze affected markets, and the incident materially reset confidence in Kelp and in cross-chain LRT bridging.

Swell (swETH and rswETH)

Swell operates both an LST (swETH) and an LRT (rswETH), bundling the two product lines under a single brand and tokenomics structure. SWELL token, Swell Chain (an L2 dedicated to restaking-aligned applications), and an active developer push give Swell a distinct ecosystem-build identity vs the more pure-play LRT competitors. Combined swETH + rswETH TVL is in the $500M-1B range.

Puffer Finance (pufETH)

Puffer's differentiation is anti-slashing technology — Secure Signer, a hardware-validated key custody system intended to prevent operator double-signing — and a focus on solo-staker decentralization. Puffer was a later entrant in 2024 with strong technical positioning. pufETH TVL typically sits at $300-700M.

Mellow Finance

Mellow operates as a modular LRT factory: third parties create curated LRT mandates with specific AVS-basket strategies, while Mellow provides the underlying smart-contract infrastructure. This positions Mellow as the issuance layer underneath bespoke LRT projects — useful for institutional buyers wanting tailored restaking exposure rather than a one-size-fits-all LRT. Mellow TVL across all curated mandates ranges $300M-1B.

The points-era economics

The 2024 stack

At peak in early 2024, an LRT yield decomposition looked something like:

In expected-value terms, leveraged YT positions on LRTs implied annualized yields of 30-100+ percent during peak speculation. This drove combined LRT TVL from $1-2B in late 2023 to $13B+ in mid-2024.

The wind-down

EIGEN launched in October 2024 with a published initial circulating value and points-conversion ratio. The mystery premium that drove YT pricing evaporated overnight. LRT-issuer tokens (ETHFI launched March 2024, REZ April 2024, KEP later 2024-2025) similarly clarified their distribution — and the leveraged points trade lost its asymmetry.

By 2025-2026, LRT yield decomposition has rationalized:

Still attractive vs base ETH staking but no longer the asymmetric leverage trade of the points era. Pendle YT-LRT markets have compressed correspondingly.

AVS quality and yield

AVSs are the demand side of the restaking equation. By April 2026 the AVS landscape includes:

Total AVS revenue paid to restakers through 2025 was lower than initial marketing suggested — supply of restaked capital materially exceeded paying AVS demand. AVS yield uplift over base staking has averaged 100-200bp rather than the 500-1000bp some early projections implied.

LRT issuers differentiate partially on AVS curation — selecting higher-quality AVSs with sustainable revenue and credible operational risk profiles, while diversifying to limit slashing concentration. This is a real but modest yield differentiator across LRTs.

Structural risks

Slashing concentration. If an LRT delegates heavily to a small set of AVSs, an operational failure or slashing event in one AVS can produce meaningful loss. Aggregated LRTs make these decisions on behalf of holders; LRT issuers' AVS curation discipline matters.

AVS quality. Many AVSs are early-stage. Protocol bugs, operational failures, or governance compromises in an AVS can cascade into restaker losses. The cryptoeconomic security model assumes correct slashing; bugs in slashing logic can produce false slashes or missed slashes.

Issuer-level risk. LRT issuers control validator keys, AVS delegations, and contract upgrade paths. A compromise or governance failure at the issuer level affects all LRT holders.

Liquidity risk. LRTs trade against ETH on AMMs. Thin liquidity during stress produces depegs — the April 2024 ezETH event being the clearest example. LRTs used as leveraged collateral on lending protocols are particularly vulnerable.

Smart contract risk. LRT contract → EigenLayer contract → AVS contracts → validator infrastructure. Multiple contract layers compound surface area. Audits help; novel exploits remain possible.

Regulatory risk. Restaking arrangements may attract securities-law scrutiny in some jurisdictions. LRT issuers operate primarily in offshore-friendly jurisdictions but the regulatory perimeter is unsettled.

What survives 2026

The LRT market consolidated meaningfully through 2024-2025. The top three by TVL have been ether.fi, Renzo and Kelp, together historically around 70 percent of category TVL — though Kelp's standing was shaken by the April 2026 $292M cross-chain bridge exploit. Smaller LRTs from the 2024 cohort have largely consolidated, pivoted, or gone dormant.

Looking forward to late 2026 and 2027, the survivors will be issuers that:

  1. Demonstrate consistent yield uplift over base staking through good AVS curation.
  2. Maintain trustworthy operator and governance practices, with public AVS delegation policies and credible audit trails.
  3. Integrate deeply with leading lending protocols (Aave, Morpho) as collateral types — this provides recurring demand and locks in DeFi composability.
  4. Extend franchises to additional restaking platforms (Symbiotic, Karak) as the restaking landscape diversifies beyond EigenLayer.
  5. Develop AVS-aligned services — running the AVSs themselves, partnering with AVS issuers for early access, providing infrastructure that AVSs need.

ether.fi and Renzo are the highest-confidence survivors — large TVL, deep integrations, active governance, multiple revenue lines. Kelp — despite its scale — faces a rebuilding of trust after the April 2026 $292M bridge exploit, while Swell and Puffer occupy clear niches likely sustained over the medium term. Mellow's modular factory model gives it a structural position that doesn't directly compete with monolithic LRTs.

Expect further consolidation over the next 12-18 months as smaller players merge or wind down, and as AVS revenue economics solidify.

Bottom line

LRTs are a durable and meaningful product category in DeFi — a layer above LSTs that captures restaking yield uplift, packaged into composable tokens. The points-era hype overstated the steady-state economics; the post-points reality is a 4-7 percent yield product with material additional risk surface compared to plain LSTs. For users who understand and underwrite the AVS, slashing, and liquidity risks, top-tier LRTs offer 100-300bp yield uplift over base staking with reasonable operational quality. For users who do not, plain LSTs (stETH, wstETH) remain the better choice.

The category through 2026 is consolidating, not expanding — a smaller number of larger, better-capitalized issuers competing on AVS curation, integration depth, and operator quality. The Cambrian-explosion phase is over; the institutional-quality phase is beginning.

Sources and further reading

About the author

DeFi Intel Research is the in-house research team at DeFi Intel, focused on on-chain capital markets, MEV, ZK infrastructure, and verifiable AI.

Last updated: 2026-07-15

Entities mentioned

ether.fiRenzoKelp DAOSwellPuffer FinanceEigenLayerEigenCloudSanctum Validator LSTs

Frequently asked questions

What is a liquid restaking token in plain English?
A liquid restaking token (LRT) is a token you receive when you deposit ETH (or a liquid staking token like stETH or wstETH) into a restaking protocol that then deposits that capital into EigenLayer. The LRT represents your share of the underlying restaked position and accrues three yield streams: (1) base ETH staking yield (~3 percent), (2) EigenLayer restaking rewards from Actively Validated Services (AVSs) using the restaked capital for security, and (3) the LRT issuer's own loyalty points or token allocations. LRTs are tradeable, usable as collateral, and composable across DeFi — exactly the relationship liquid staking tokens (LSTs) have to base staking, but one layer higher in the staking stack.
Why did LRTs explode in 2024 and what is happening now?
From late 2023 through mid-2024, LRTs caught a perfect-storm narrative: EigenLayer's points program promised material future EIGEN airdrops, LRT issuers ran their own points programs, and Pendle let users leverage YT for amplified points exposure. eETH, ezETH, rsETH, swETH, and pufETH each ran promotional points campaigns. Combined LRT TVL crossed $13B at peak in mid-2024. Following EIGEN's launch and points-conversion clarification, LRT TVL fell to $7-9B steady state by 2025-2026 as the leveraged points trade unwound. The LRTs that survived have transitioned to genuine yield products competing on AVS quality and integration depth rather than points marketing.
How is each major LRT different?
Six players matter. (1) ether.fi (eETH/weETH) — largest LRT by TVL ($3-4B in 2026), distinguishes itself with a delegated-validator model where stakers run their own validators, ETHFI token, and broad DeFi integrations. (2) Renzo (ezETH) — second-largest, focused on multi-chain availability and aggressive points distribution; faced a depeg event in April 2024 that taught the market about AVS-curve liquidity risk. (3) Kelp DAO (rsETH) — accepts LSTs as deposits (stETH, ETHx, sfrxETH) rather than only ETH, broader collateral acceptance. (4) Swell (rswETH) — combines LST (swETH) and LRT (rswETH) franchises under one issuer with native L2 (Swell Chain). (5) Puffer Finance (pufETH) — anti-slashing tech (Secure Signer hardware-validated keys), focus on solo-staker decentralization. (6) Mellow Finance — modular LRT factory letting third parties create curated LRT mandates with specific AVS basket strategies.
What are AVSs and why do they matter for LRT yield?
AVSs (Actively Validated Services) are protocols that pay EigenLayer restakers for cryptoeconomic security. They include data availability layers (EigenDA), oracle networks, bridges, MEV-related services, ZK provers, and various middleware. An AVS rents security from EigenLayer's restaked capital pool — restakers signal which AVSs they delegate to and earn yield from those AVSs. By April 2026 there are roughly 25-40 live AVSs with varying revenue models. AVS yield in 2024-2025 was lower than initially marketed (often 1-3 percent annualized rather than 5-10 percent), partly because AVSs are early-stage and partly because supply of restaked capital exceeded paying AVS demand. Higher-quality AVS picks differentiate LRTs that actively manage delegation, but yield uplift over base staking remains modest.
What were the points-era economics and how have they changed?
In the 2024 points era, LRT yield decomposition looked roughly like: 3 percent ETH staking + 1-3 percent AVS rewards (often paid in points or TBD tokens) + speculative value of EigenLayer points (priced at $0.05-0.15 per point on secondary markets) + LRT issuer's own loyalty points (additional speculative value) = effective 15-35 percent annualized in expected-value terms during peak speculation. Post-EIGEN launch, the speculative components rationalized: EIGEN price set the points-conversion ratio, LRT loyalty programs converted into governance tokens (ETHFI, REZ, KEP) on actual schedules, and the YT trade compressed. Effective LRT yield by 2026 sits at 4-7 percent annualized — base staking plus a modest AVS uplift plus residual loyalty distributions. Still attractive vs base ETH staking, but no longer the asymmetric trade of the points era.
What are the structural risks of holding an LRT?
Five categories. (1) Slashing concentration — if an LRT delegates heavily to a single AVS that experiences a slashing event, the LRT can take meaningful losses. Aggregated LRTs make individual AVS risk decisions on behalf of holders. (2) AVS quality — many AVSs are early-stage; protocol bugs or operational failures in an AVS can cascade into restaker losses. (3) Issuer key management — LRT issuers control critical operator keys and validator infrastructure; an issuer-level compromise affects all holders. (4) Liquidity risk — LRTs trade against ETH on AMMs (Curve, Balancer, Uniswap V3); thin liquidity during stress can produce material depegs (the Renzo ezETH depeg in April 2024 saw ezETH fall sharply below peg before recovering). (5) Smart contract risk — multiple contract layers (LRT contract, EigenLayer contract, AVS contracts) compound surface area. The LRT is only as safe as the weakest contract in the chain. Cross-chain bridge risk is now a proven category: Kelp's LayerZero bridge was exploited for ~$292M in April 2026 via a single-verifier misconfiguration.
Are LRTs better than just holding ETH or stETH?
It depends on the user. For pure ETH yield maximization with low operational complexity, stETH or wstETH from Lido remains the simplest answer — one yield stream, deepest liquidity, well-understood risk. For users wanting to capture restaking yield uplift and willing to underwrite the additional risk surface (slashing concentration, AVS quality, issuer governance), an LRT from a top-tier issuer (ether.fi, Renzo) is a reasonable choice — expect 100-300bp yield uplift over base staking with materially higher tail risk. For users explicitly betting on AVS yield growth or who hold strong views on which LRT will accrue the most value capture as restaking matures, LRT exposure is the natural play. For users who do not understand what an AVS is and what slashing concentration means, LRTs are not appropriate.
Which LRTs survive 2026 and which consolidate?
The market has already consolidated meaningfully. The top three (ether.fi, Renzo, Kelp) hold ~70 percent of LRT TVL, though Kelp's standing was jolted by the April 2026 $292M LayerZero-bridge exploit — the biggest DeFi hack of 2026 and a reminder that scale does not equal safety. Swell and Puffer occupy clear niches (LST+LRT under one roof for Swell, anti-slashing tech for Puffer). Mellow has carved out a structural position as the modular factory layer underneath bespoke LRT issuers. Smaller LRT projects from 2024 — many launched on the back of the points-era hype — have either consolidated, pivoted, or wound down. Looking forward to late 2026 and beyond, the survivors will be those that (1) demonstrate consistent yield uplift over base staking, (2) maintain trustworthy operator and AVS-delegation governance, (3) integrate deeply with leading lending protocols (Aave, Morpho) as collateral, and (4) extend their franchises into new restaking layers (Symbiotic, Karak) and AVS-aligned services. ether.fi and Renzo are the highest-confidence survivors; the second tier consolidates further over the next 12-18 months.

Related entities

Related guides

Stay current on this topic

Get the weekly DeFi Intel brief — entity-graph intelligence delivered weekly, free to your inbox.

Subscribe free See Pro →

Frequently asked questions

What are liquid restaking tokens (LRTs) and how do they work?

LRTs are tokens minted when a user deposits ETH or an LST into an LRT issuer's contract, which stakes the ETH, restakes the resulting position into EigenLayer, and mints a tradeable token representing the restaked position that accrues base staking yield, AVS rewards, and any issuer points.

Which LRT issuer was the largest by TVL through 2024-2026?

ether.fi (eETH/weETH) was the largest LRT by TVL through 2024-2026, typically managing $3-4B.

What distinguishes ether.fi's validator model from other LRT issuers?

ether.fi uses a delegated-validator model where stakers can opt to run their own validator infrastructure under ether.fi's stack, giving it a solo-staker-friendly story and a more decentralized operator profile than Lido.