Liquid Restaking 2026: Complete Guide (EigenLayer, LRTs)

Liquid restaking is the most important DeFi primitive of the 2024-2026 cycle. By stacking restaking on top of regular liquid staking, the category unlocked tens of billions of dollars of additional yield on Ethereum and extended the model to Bitcoin, Cosmos, and beyond. EigenLayer kicked off the wave; Renzo, Kelp, Puffer, and the rest built the consumer-facing LRT products; Symbiotic and Karak created the second-generation protocols. This complete 2026 guide breaks down how liquid restaking actually works, the major protocols compared, the yields and risks, and where the category goes next.

We will cover the foundations of restaking, walk through EigenLayer and the AVS layer, deep dive each major LRT (Renzo ezETH, Kelp rsETH, Puffer pufETH, YieldNest ynd), look at the universal restaking protocols (Symbiotic, Karak), cover the Bitcoin restaking expansion (BounceBit), compare yields and risks, and close with an extensive FAQ.

What is restaking, and what makes it "liquid"?

Restaking is the practice of taking already-staked ETH (or equivalent staked tokens on other chains) and using it as cryptoeconomic security for additional services beyond the base chain. The mechanic is straightforward: a validator who has 32 ETH staked on Ethereum can opt in to provide that same stake as security to a third-party Actively Validated Service (AVS), accepting the additional slashing risk in exchange for additional rewards.

Why does anyone want this? Because building new infrastructure (oracles, rollups, bridges, data availability layers) requires cryptoeconomic security. Building that security from scratch (recruiting validators, bootstrapping a token, accumulating stake) is expensive and slow. Restaking lets new services rent security from existing validators, paying them in tokens or fees for the additional slashing exposure. EigenLayer pioneered this model on Ethereum in 2023, and it has since extended to every major staking ecosystem.

Liquid restaking takes the next step. Instead of users locking their ETH directly into the restaking protocol (which would be illiquid), Liquid Restaking Token (LRT) issuers like Renzo, Kelp, Puffer, and YieldNest accept ETH deposits, restake it on behalf of users, and issue a liquid token (ezETH, rsETH, pufETH, ynLSDe) that represents the user's claim. The LRT can be used in DeFi: traded, used as collateral, deposited in yield aggregators. The user keeps liquidity while earning the restaking yield.

For the foundational deep dive on what an LRT is and how it differs from a regular liquid staking token, see What is a Liquid Restaking Token (LRT) DeFi guide 2026. For the broader explainer on restaking as a shared security strategy, see Liquid restaking explained: the shared security strategy.

EigenLayer: the foundation

EigenLayer is the original and still-largest restaking protocol on Ethereum. Founded by Sreeram Kannan at the University of Washington, EigenLayer launched mainnet in mid-2023 and grew to over $15B in total value locked (TVL) by 2024-2025 before partial rotation to competing protocols in 2026.

The architecture has three layers. First, users deposit ETH (either natively staked or liquid-staked, like stETH or rETH) into EigenLayer. Second, the deposited ETH provides additional security to AVSs (Actively Validated Services) that opt in to EigenLayer's security pool. Third, the AVSs pay rewards in their own tokens or in ETH, which flow back to the underlying restakers and to operators (the entities running validator infrastructure for AVSs).

The major AVSs on EigenLayer include EigenDA (Eigen's own data availability layer), Lagrange (zk coprocessor), Witness (a verifiable timestamp service), AltLayer (rollup-as-a-service), and dozens of smaller projects. Each AVS has its own slashing rules and reward structure. Restakers can opt in to specific AVSs or accept the default basket curated by their LRT issuer.

Deep dives on the protocol: What is EigenLayer restaking protocol guide 2026 and the original What is EigenLayer restaking explained. For the comprehensive view across the whole category including EtherFi, see What is restaking: EigenLayer + EtherFi complete guide 2026.

The major LRTs

Renzo (REZ) and ezETH

Renzo is one of the largest LRT issuers on EigenLayer, with ezETH as the user-facing token. The protocol accepts ETH deposits, restakes the ETH through EigenLayer with a curated basket of AVSs, and issues ezETH to depositors. The REZ token coordinates governance and protocol incentives.

Renzo's value proposition is curation: rather than forcing users to choose which AVSs to opt into, Renzo runs a curated allocation strategy that aims to maximize risk-adjusted yield. The trade-off is that users delegate the AVS-selection decision to Renzo's team. Deep dive: What is Renzo REZ ezETH liquid restaking EigenLayer guide 2026.

Kelp DAO and rsETH

Kelp DAO is another major LRT issuer with rsETH as the user-facing token. The architecture is similar to Renzo (deposit ETH, get rsETH, earn restaking yield), with differences in AVS curation, fee structure, and integration with the broader DeFi ecosystem. Kelp has been particularly active in Layer 2 integrations, with rsETH available on Arbitrum, Optimism, Base, and other L2s.

Deep dive: What is KelpDAO rsETH liquid restaking EigenLayer guide 2026.

Puffer Finance and pufETH

Puffer Finance differentiates with its UniFi anti-slashing technology and its native validator architecture. Rather than relying entirely on third-party node operators, Puffer enables permissionless validator participation with anti-slashing guarantees that aim to reduce the risk of validator misbehavior. The PUFFER token coordinates governance.

Deep dive: What is Puffer Finance pufETH liquid restaking guide 2026.

YieldNest and ynLSDe / YND

YieldNest takes the "max LRT" approach: instead of restaking through one underlying protocol, ynLSDe and YieldNest's other LRTs aggregate yields across multiple restaking protocols, including EigenLayer, Symbiotic, and Karak. The architecture lets users access broader restaking exposure through a single token, at the cost of additional smart-contract complexity.

Deep dive: What is YieldNest YND liquid restaking max LRT guide 2026.

Liquid Restaking Tokens compared (2026)

Universal restaking: Symbiotic and Karak

Symbiotic

Symbiotic launched in 2024 as a second-generation universal restaking protocol. Rather than restricting users to staking ETH and serving Ethereum-based AVSs, Symbiotic supports any asset and any consumer network. Users can stake ETH, LSTs, LRTs, or other tokens, and networks can opt in to use any of these assets for security. The flexibility is the value proposition; the trade-off is that the security guarantees are more complex to reason about.

Deep dives: What is Symbiotic Protocol: permissionless restaking guide 2026 and the original What is Symbiotic: permissionless restaking guide.

Karak Network

Karak is another universal restaking protocol, with a focus on speed of integration and a broader asset acceptance list (including BTC-pegged assets and stablecoins, not just ETH). Karak has built its own L2 (Karak K2) to host the restaking activity, which gives it more control over the throughput and finality characteristics than protocols deployed directly on Ethereum mainnet.

Deep dives: What is Karak Network universal restaking guide 2026 and What is Karak: universal restaking explained.

Bitcoin restaking: BounceBit and the BTC layer

Restaking is no longer Ethereum-only. BounceBit pioneered the Bitcoin restaking thesis with a CeDeFi architecture that lets BTC holders earn restaking-style yield without giving up custody of their underlying Bitcoin. The architecture uses centralized custodians (Mainnet Digital, Ceffu) to hold the BTC while issuing a wrapped representation (bbtc) that can be used in the BounceBit chain's restaking and DeFi economy.

The BB token coordinates the protocol economy. Validators stake BB to participate in consensus; restakers earn yield from the underlying CeFi BTC strategy plus on-chain DeFi activity. Deep dive: What is BounceBit BB CeDeFi Bitcoin restaking chain guide 2026.

The broader top-5 restaking protocol comparison (covering EigenLayer and the major alternatives) is in Top 5 restaking protocols on EigenLayer 2026.

Yield sources: where the APY actually comes from

A liquid restaking yield is the sum of several distinct components, and understanding which component contributes how much is crucial for assessing risk-adjusted returns.

Base staking yield (3-5% APR). The underlying Ethereum staking yield from running a validator. This is the safest component and the one that has the longest track record. It is denominated in ETH and paid by the protocol via block rewards and priority tips.

AVS rewards (variable, 0.5-5% effective APR). Payments from AVSs to restakers for providing additional security. Denominated in the AVS's own token (or sometimes ETH), with the actual yield depending on the AVS's tokenomics and the size of the restaker pool.

Protocol airdrop points (highly variable). Most LRT issuers and restaking protocols run points programs that accrue to depositors and may convert to token airdrops at a future date. EigenLayer points, Karak XP, Symbiotic points, and individual LRT points all have value (in expectation) but the actual realized value depends on the token launch terms.

DeFi integration yields. LRTs can be used as collateral in lending markets, deposited in yield aggregators, and provided as liquidity in AMM pools, each of which adds an additional yield layer on top of the underlying restaking. These layered yields amplify returns but also amplify smart-contract risk.

The broader landscape of Ethereum yield options (native staking, liquid staking, restaking, LRTs) is covered in Ethereum yield: liquid staking and restaking 2026. For a decision-tree comparison of when to use which approach, see Native staking vs. liquid staking vs. restaking decision tree.

Risks: where things can go wrong

The April 2024 ezETH depeg: case study

The most-studied stress event in the LRT category was the April 24, 2024 ezETH depeg. After Renzo announced the REZ tokenomics with a smaller-than-expected allocation to ezETH holders, secondary market sellers rushed to exit, while the protocol's native withdrawal queue was insufficient to absorb the flow. The result was an ezETH price drop from approximately 1.0 ETH to 0.93 ETH (a 7% depeg) over several hours, with cascading liquidations in lending markets where ezETH was used as collateral.

The event was instructive for the whole category. First, it demonstrated that LRTs can depeg during stress even when the underlying restaking is performing normally. Second, it showed that token-launch events are a specific risk window: airdrop expectations affect demand, and demand affects the peg. Third, it highlighted the importance of withdrawal queue capacity and secondary market depth as components of risk management.

In the aftermath, most LRT issuers improved their withdrawal queue mechanics and added stress-event procedures. The category recovered, but the event reset expectations about LRT safety. The lesson: LRT yields are real, but the wrapper introduces meaningful additional risk that does not exist with native staking.

Choosing an LRT: practical considerations

For users who decide that liquid restaking fits their portfolio, the choice of which LRT to use comes down to several practical factors.

1. AVS curation philosophy

Each LRT issuer makes different AVS selection choices. Some are conservative (a small basket of established AVSs), some are aggressive (broader exposure to newer AVSs for higher yield). Match the LRT to your own risk tolerance.

2. DeFi integration breadth

If you plan to use the LRT as collateral or in yield strategies, check which protocols accept it. The largest LRTs (weETH, ezETH, rsETH) have the broadest DeFi integration. Smaller LRTs may have fewer integration options.

3. Withdrawal queue characteristics

Check the native withdrawal queue mechanics. How long is the standard unbonding? Is there a fast-exit mechanism (often via a third-party AMM)? What happens during stress events? The withdrawal queue is the user's emergency exit, and its design matters more during the moment you actually need it.

4. Track record and team

LRT issuers with longer track records and known teams (EtherFi, Renzo, Kelp, Puffer) have lower implementation-risk than newer entrants. The protocol can be technically sound and still have operational issues that newer teams are more likely to encounter.

5. Fee structure

Each LRT charges a management fee, usually 10% to 20% of yield. Compare the net yield (after fees) rather than the headline yield. The difference between a 12% gross yield with 20% fees (9.6% net) and a 11% gross yield with 10% fees (9.9% net) is meaningful over multi-year periods.

Looping and leveraged LRT strategies

A common advanced strategy in the LRT ecosystem is looping: deposit an LRT (like weETH or ezETH) as collateral in a lending market, borrow stablecoins or ETH against it, use the borrowed funds to buy more of the LRT, and deposit that into the collateral pool again. Repeated several times, the loop amplifies the underlying restaking yield by 2x to 5x, depending on the loan-to-value ratio and the cost of borrowing.

The math works because the LRT yield (6-15%) typically exceeds the borrow rate on Aave, Morpho, or Spark (3-7% for ETH, 4-10% for stablecoins). The spread is the leveraged yield. The risks are also amplified: a temporary depeg in the LRT (like the April 2024 ezETH event) can trigger cascading liquidations across the looped position, wiping out months of yield gains in a single bad afternoon.

For practitioners, the rule of thumb is that looping is appropriate when the spread is wide enough to absorb significant LRT depeg risk and the user has the discipline to unwind the loop quickly if conditions change. Looping during stress events (when depegs are most likely) is the surest way to blow up the position. Looping during quiet markets is one of the highest yields available in DeFi for sophisticated users.

The AVS landscape: who is paying for security

For restaking to have durable value, AVSs need to be real businesses paying real fees for real services. The 2025-2026 cycle saw a transition from points-and-airdrops AVS economics to early-stage fee-paying AVS economics, with several projects starting to generate ETH-denominated fees that flow back to restakers.

EigenDA is Eigen's own data availability layer, competing with Celestia and EigenLayer-native projects. EigenDA generates fees from rollups paying for data availability and has been one of the most successful AVS launches in terms of fee revenue.

Lagrange is a zk coprocessor that provides verifiable computation services to applications. Lagrange has generated meaningful AVS rewards through 2025-2026 and is one of the prominent examples of AVS technology with real demand.

Witness is a verifiable timestamp service that uses restaked ETH to provide tamper-proof timestamping for legal and compliance applications. The use case is narrower than EigenDA or Lagrange but the AVS revenue model is clearer.

AltLayer provides rollup-as-a-service infrastructure with EigenLayer security. AltLayer was one of the first AVSs to launch and remains active, though it has not generated the AVS revenue some early projections suggested.

The wider universe of AVSs includes dozens of projects in various stages of development, with the pattern that most are still in the points-and-promise phase rather than the fee-revenue phase. The 2026-2027 cycle will determine which AVSs convert points programs to durable fee revenue and which fade as the points cycle ends.

Where liquid restaking goes next

Three trends will define the liquid restaking category through 2026-2027. First, consolidation: the category has dozens of LRT issuers competing for similar capital, and the smaller issuers will either find a niche or get squeezed out. The top 4-5 LRTs (weETH, ezETH, rsETH, pufETH, ynLSDe) will likely capture the majority of long-term TVL. Second, multi-asset expansion: BTC restaking (BounceBit), SOL restaking, and stablecoin restaking are all areas of active development. The total addressable market for restaking grows substantially if it extends beyond ETH. Third, AVS commercialization: the AVS layer has been points-and-airdrops driven so far. The transition to real fee revenue (paid in ETH or stablecoins for actual services) is the long-term test of whether restaking has durable economic value or whether it was a 2024-2025 cycle phenomenon.

The bigger picture: liquid restaking is one of the few categories in DeFi that delivered on its initial thesis in 2024-2025. Tens of billions of dollars of cryptoeconomic security have been provisioned for new infrastructure, and the LRT wrapper has made it accessible to retail. The challenge for 2026-2027 is moving from a points-driven economy to a fee-driven economy without losing the demand that points programs unlocked.

FAQ: Liquid Restaking 2026

Tax treatment of LRTs and restaking yield

Tax treatment of LRTs in the US is still developing in 2026, and treatment varies by jurisdiction. The general framework is that LRT yields are likely treated as ordinary income at the moment they accrue, similar to staking rewards. This is consistent with the IRS guidance that staking rewards are income at the time the rewards become available to the staker, even if not yet sold.

The complication for LRTs is that the rewards accrue continuously to the LRT price (the LRT becomes worth incrementally more ETH over time, rather than receiving separate reward tokens). Some tax software treats this as deferred until the LRT is sold; others treat each "rebase" as income. Tax professionals working with crypto have generally taken the latter view, but the IRS has not issued LRT-specific guidance.

Points programs and airdrops are a separate tax issue. Points themselves are typically not taxable until they convert to tokens, at which point the fair market value of the airdrop is treated as ordinary income. The longer the gap between points accrual and token launch, the more complex the basis tracking becomes.

For high-net-worth restakers, working with a crypto-specialist tax professional is essentially required. The combination of multiple LRT yields, looped positions, multiple AVS reward tokens, and multiple points programs can produce hundreds of taxable events per year that are not handled cleanly by general-purpose tax software.

EtherFi as the LRT category leader

No discussion of liquid restaking is complete without a focused look at EtherFi, the LRT issuer that has captured the largest market share by TVL through 2024-2026. EtherFi launched in early 2023 as a non-custodial Ethereum staking protocol with a focus on keeping users in control of their withdrawal keys. The product evolved into weETH, the wrapped version of EtherFi's eETH, which became the dominant LRT in the EigenLayer ecosystem.

EtherFi's positioning rests on three pillars. First, the non-custodial architecture: users keep their withdrawal credentials, which differentiates EtherFi from most competing protocols where the issuer controls withdrawals. Second, the breadth of DeFi integration: weETH is accepted as collateral on virtually every major Ethereum DeFi protocol, which makes it the most flexible LRT for advanced strategies. Third, the multi-protocol restaking: EtherFi has expanded beyond pure EigenLayer to also include Symbiotic, which gives weETH holders exposure to multiple AVS ecosystems.

The ETHFI token coordinates EtherFi governance and protocol incentives. The token launched in 2024 and has gone through multiple market cycles since. For the full EtherFi deep dive in the broader restaking context, see What is restaking: EigenLayer + EtherFi complete guide 2026.

Restaking vs. native staking: when to use which

Liquid restaking is not the right choice for every ETH holder. The decision between native staking, regular liquid staking, and liquid restaking depends on the user's risk tolerance, capital size, and DeFi sophistication.

Native staking (running your own validator). Best for users with at least 32 ETH, technical comfort with validator operation, and a long-time horizon. Yields are pure base staking (3-5%), with no smart-contract risk beyond the Ethereum protocol itself. Slashing risk is real but limited to misbehavior of your specific validator.

Liquid staking (stETH, rETH). Best for users who want liquid exposure to ETH staking yield without running their own validator. Lower risk than LRTs because no restaking layer is added. Lower yield because no AVS rewards. Lido stETH, Rocket Pool rETH, and similar tokens dominate this layer.

Liquid restaking (LRTs). Best for users seeking higher yield who are comfortable with the multi-layer smart-contract risk. The additional yield over plain liquid staking comes with proportionate additional risk. Most appropriate as a moderate-to-large allocation for sophisticated DeFi users, not as a sole ETH-yield strategy.

The decision tree comparison: Native staking vs. liquid staking vs. restaking decision tree.

A walkthrough: depositing into an LRT

For users who want a practical sense of what depositing into an LRT actually looks like, here is the step-by-step process for a typical Renzo ezETH deposit. The mechanics are similar across other major LRT issuers (Kelp, Puffer, EtherFi).

Step 1: Connect wallet. Go to the Renzo app at renzoprotocol.com (verify the URL carefully; phishing sites are common in this category). Connect MetaMask, Rabby, or another supported wallet. Confirm you are on Ethereum mainnet.

Step 2: Approve and deposit ETH. Enter the amount of ETH (or wstETH, or another supported asset) you want to deposit. The first transaction approves the Renzo smart contract to spend your tokens; the second deposits the funds. Gas costs vary with network conditions but typically total $5 to $30 for the full deposit flow.

Step 3: Receive ezETH. The protocol issues ezETH 1:1 with deposited ETH (subject to small price differences in secondary markets). The ezETH balance appears in your wallet immediately. From this point, the deposit is earning the underlying restaking yield, plus EigenLayer points, plus Renzo's own points program.

Step 4 (optional): use ezETH in DeFi. Deposit ezETH in Aave, Morpho, Spark, or another lending market to use as collateral. Add liquidity to an ezETH/ETH AMM pool to earn additional fees. Use ezETH in Pendle to lock in fixed yield or trade speculation on future yield. Each integration adds yield with proportionate risk.

Step 5: monitor and exit. Track the underlying performance via DeFiLlama or Renzo's own dashboard. When you want to exit, either redeem ezETH through the native withdrawal queue (which typically takes 7 days due to Ethereum unbonding) or swap ezETH for ETH on a secondary market AMM (Uniswap, Curve, Balancer) for immediate liquidity but at the prevailing market price.

Bottom line

Liquid restaking grew from zero to tens of billions of dollars of TVL in less than two years, delivering on the original EigenLayer thesis of cryptoeconomic security as a service. EtherFi, Renzo, Kelp, Puffer, and YieldNest dominate the LRT layer; Symbiotic and Karak provide the second-generation universal restaking infrastructure; BounceBit extends the model to Bitcoin. The yields are real, the risks are stacked, and the 2026-2027 cycle will test whether the AVS economy can transition from points-driven to fee-driven sustainability. Use the satellite guides linked throughout this pillar to deep dive into each protocol and LRT, and remember that liquid restaking is one of the more risk-layered products in DeFi, with returns that compensate the underlying complexity.

Related Guides

• What Is Renzo (REZ)? ezETH Liquid Restaking on EigenLayer Guide 2026
• What Is KelpDAO (rsETH)? Liquid Restaking on EigenLayer Guide 2026
• What Is Restaking: Complete EigenLayer & EtherFi Guide (2026)
• What Is EigenLayer: Complete Restaking Protocol Guide (2026)
• What is EigenLayer? Restaking Explained