Restaking and Liquid Restaking Tokens
Executive summary
Restaking, the practice of repurposing staked ETH (or other staked assets) to secure additional protocols, exploded into the largest DeFi category between 2023 and 2025 before settling into a more measured 2026 equilibrium. EigenLayer crossed $20B+ TVL at peak in mid-2024 before normalizing as the points era ended and the EIGEN token launched. A constellation of liquid restaking token issuers (ether.fi, Renzo, Kelp, Swell, Puffer) packaged restaked positions into tradable LRTs that proliferated across DeFi as collateral. By 2026, the category is bifurcating: native ETH restaking on EigenLayer remains the institutional core, alternative platforms (Symbiotic, Karak, Babylon for BTC) compete for differentiated AVS demand, and the AVS market itself is transitioning from a points-driven launch phase into a real-yield phase where slashing risk and AVS economics actually matter.
Origin and mechanism: what restaking actually does
Restaking, conceived by EigenLabs in 2021 and launched in beta in 2023, repurposes the cryptoeconomic security of Ethereum staking by allowing stakers to opt into additional slashing conditions imposed by external protocols (Actively Validated Services, or AVSs). The mechanism is straightforward: a staker either delegates their native validator (via withdrawal credentials) or deposits an LST or pure ETH into the EigenLayer contracts, then opts into one or more AVSs. If the staker violates an AVS's rules, a portion of their stake is slashed and either burned or redirected. In return, the staker earns AVS rewards on top of base ETH staking yield. The thesis is that any new protocol needing economic security (a sequencer, a data-availability layer, a coprocessor, an oracle network, a bridge) can rent it from EigenLayer's stake pool rather than bootstrap its own token and validator set. The cost is that slashing risk compounds: a single staker exposed to ten AVSs faces ten distinct slashing surfaces. By 2026, native ETH restaking is the dominant stake type, and AVSs began enforcing real slashing conditions in 2025 after EigenLayer enabled slashing on mainnet.
EigenLayer and the AVS economy
EigenLayer is the dominant restaking platform with at-peak $20B+ TVL in 2024, normalizing to roughly $12-15B by Q1 2026 as some points-driven capital exited and slashing-aware capital entered. The AVS roster grew from a handful of high-profile launches in 2024 (EigenDA, AltLayer, Lagrange, Witness Chain) to dozens of live AVSs by 2026, spanning data availability, coprocessing, fast finality, decentralized sequencers, oracle networks, and bridge security. EigenDA, EigenLayer's own data availability service, became the largest AVS by stake and a credible alternative to Celestia for L2 data availability, integrated by several Optimism Superchain rollups and a growing list of L3s. The AVS reward landscape in 2026 looks different from 2024: rather than purely native token rewards, AVSs increasingly pay in stablecoins or restaked ETH, anchoring real-yield economics. Slashing went live on EigenLayer mainnet in 2025, enabling operator-level slashing across AVSs after repeated delays during the preceding testnet period. The EIGEN token launched in October 2024 as a forkable intersubjective token meant to slash for behaviors not provable onchain (such as DA layer corruption), and underwent its own reward and governance phase through 2025-2026.
Liquid Restaking Tokens (LRTs) and the issuer ecosystem
Native restaking carries the friction of locked positions and direct AVS selection. Liquid Restaking Tokens solved that by issuing a tradable claim on a managed restaking position, much as Lido's stETH solved the analogous problem for staking. By 2026, the LRT issuer landscape consolidated around five main players. ether.fi, the largest LRT issuer, crossed $7-8B TVL at peak with eETH/weETH and developed a complementary product stack (cash card, restaking-collateralized credit). Renzo's ezETH crossed $3-4B before suffering a notable depeg event in April 2024 that became a cautionary tale; it recovered structurally and remains a top-three LRT. Kelp DAO's rsETH grew steadily, focusing on diversified AVS exposure. Swell's rswETH wrapped its existing swETH product into a restaking layer. Puffer Finance's pufETH took a more anti-slashing-focused engineering approach. Each LRT had distinct AVS-selection policies, fee structures, and risk frameworks. By 2026 LRTs are widely accepted as collateral on Aave, Morpho, Spark, Pendle, and most major lending venues, with Pendle in particular dominating the LRT yield-tokenization market. Aggregate LRT TVL settled in the $10-15B range, down from points-era peaks but stable on a real-yield basis.
Beyond Ethereum: Babylon, Karak, Symbiotic, and BTC restaking
Restaking quickly became a multi-platform category. Babylon, on Bitcoin, pioneered native BTC restaking using BTC's timelock script to provide cryptoeconomic security to PoS chains via a slashable signature commitment. Babylon's mainnet launched in stages through 2024-2025 and crossed multi-billion-dollar BTC TVL, primarily securing Cosmos-zone PoS chains. Symbiotic, a competitor on Ethereum, built a more permissive restaking architecture that supports any ERC-20 as restakable collateral, attracting protocols that wanted bespoke economic security. Karak Network positioned itself as a multi-chain restaking layer with its own L2 (K2) for AVS execution. The competitive dynamic between these platforms is real but not zero-sum: AVSs can multi-source security across EigenLayer, Symbiotic, Karak, and Babylon, mixing ETH and BTC collateral. By 2026, the rough TVL split is roughly 60% EigenLayer (ETH-denominated), 15% Babylon (BTC-denominated), and the remainder split among Symbiotic, Karak, and emerging multi-asset restaking platforms. Solana saw early restaking experimentation through Solayer and Jito's restaking initiatives, though scale remained smaller relative to EVM equivalents.
The points era and the EIGEN launch
Restaking's growth from late 2023 through mid-2024 was inseparable from the points meta. EigenLayer opened restaking deposits in mid-2023 and accrued points to stakers; ether.fi, Renzo, Kelp, and others stacked their own points on top, creating a layered points economy that drew tens of billions in TVL on the speculative bet that points would convert into valuable token allocations. The EIGEN token's October 2024 launch and the subsequent LRT issuer airdrops (ether.fi's ETHFI in March 2024, Renzo's REZ in April 2024, Kelp's KEP, Puffer's PUFFER) materialized those points but in many cases at valuations that disappointed retail expectations. The post-airdrop period saw meaningful TVL outflow as opportunistic capital rotated to other narratives. By 2026, the LRT issuer business model evolved away from points and toward direct AVS reward distribution: holders earn from actual AVS yields rather than expected token allocations. This is healthier economically but slower-growing. The LRT issuer tokens themselves traded heavily through 2025 with mixed fundamental support, and several issuers diversified into adjacent products (ether.fi's cash and credit cards, Puffer's based rollup) to reduce single-product risk.
Slashing risk and the cryptoeconomic stack
The headline risk of restaking is slashing: if an AVS triggers a slashing event, restaked positions absorb losses. By 2026, multiple slashing events have occurred in production, though most have been small and contained. The risk is not uniform across LRTs: an LRT that opts into 15 AVSs has 15 distinct slashing surfaces, and the fattest tail risks are correlated slashing events (a software bug across multiple AVSs sharing a client implementation, or a coordinated operator failure). LRT issuers have professionalized risk management, with most publishing detailed AVS allocation policies and conducting due diligence before opting in. The role of the operator (the entity actually running the AVS software) is critical: most LRTs delegate to a curated set of operators (P2P, Kiln, Figment, Coinbase Cloud, A41) rather than running infrastructure themselves. The operator layer is itself a centralization vector if too few operators secure too many AVSs. EigenLayer, Symbiotic, and Karak each have their own slashing dispute mechanisms with varying degrees of decentralization.
Outlook through 2027
Restaking's near-term trajectory hinges on three questions. First, does AVS demand grow into the existing supply of restaked stake? In 2024-2025, restaked TVL exceeded AVS demand for security, leaving stakers under-utilized and yields compressed. By 2026 the gap began closing as more AVSs launched, but the equilibrium remains unclear. Second, do alternative restaking platforms (Symbiotic, Karak, Babylon) carve out durable niches, or does the network-effect logic concentrate around EigenLayer? Third, does the LRT market consolidate to 2-3 winners, or does a long tail persist? The most likely 2027 scenario: EigenLayer remains the dominant ETH-restaking platform with a normalized $15-25B TVL; ether.fi and Renzo dominate ETH LRTs; Babylon scales BTC restaking into the $10B+ range; Symbiotic and Karak persist as differentiated alternatives for protocols wanting permissionless or multi-asset collateral. Slashing events become routine but not catastrophic, AVS economics stabilize on stablecoin or ETH-denominated yield, and restaking-as-a-service becomes a mature B2B category rather than a retail speculation theme. The original thesis - that programmable cryptoeconomic security is a real primitive - looks vindicated; the financial-engineering excess on top has cooled.
Watch points
- EigenLayer AVS revenue trajectory and ratio of native-token vs ETH/USD-denominated rewards
- Slashing events and their treatment by major LRT issuers
- Babylon BTC TVL and the AVS roster on the BTC-restaking side
- Symbiotic/Karak share of new AVS launches relative to EigenLayer
TL;DR
Restaking matured from points-driven hype into a real-yield AVS economy by 2026, with EigenLayer dominant on ETH, Babylon scaling BTC restaking, and a stabilized LRT issuer market in the $10-15B range.
Get DeFi Intel research in your inbox
Weekly long-form coverage of papers, incidents, jurisdictions, chains, tokens and the people building them. Free tier covers headlines; Pro adds the analyst-grade breakdowns.