The Ethereum restaking sector, which peaked above $22 billion in total value locked in August 2025, has contracted to $5.10 billion on EigenLayer as of September 2026. The largest liquid restaking protocol, ether.fi, is completing a full exit from EigenLayer by the end of Q4 2026, reducing its re...
"There were no meaningful yield opportunities in restaking and there was some perceived risk from stakers, so we decided it made sense to exit." — Mike Silagadze, CEO, Ether.fi
The Ethereum restaking sector, which peaked above $22 billion in total value locked in August 2025, has contracted to $5.10 billion on EigenLayer as of September 2026. The largest liquid restaking protocol, ether.fi, is completing a full exit from EigenLayer by the end of Q4 2026, reducing its restaked assets to below 1% of its portfolio. The five largest remaining liquid restaking token (LRT) protocols — Renzo, Kelp, Swell, Puffer Finance, and Bedrock — generated a combined $953,350 in gross profit during Q2 2026, down from $2.18 million three quarters earlier.
The underlying economics tell a stark story. As of September 8, 2026, the restaking sector held $10.02 billion in TVL but generated only $99,977 in weekly fees. By comparison, liquid staking's $51.87 billion in TVL produced $27.35 million in weekly fees — roughly 53 times more revenue per dollar secured. EigenLayer itself has rebranded as EigenCloud and pivoted toward verifiable computing, with restaked collateral repositioned as a subordinate infrastructure layer. The EIGEN token has fallen 93.6% from its 2024 peak to approximately $0.26, with a market capitalization of $238 million.
Restaking emerged in 2023-2024 as an extension of Ethereum's proof-of-stake security model. The premise: staked ETH already secures Ethereum's consensus layer, but that same collateral could simultaneously secure additional services — oracle networks, bridges, data availability layers — called Actively Validated Services (AVSs). Operators would earn fees from both Ethereum staking and AVS participation, theoretically generating layered yields.
EigenLayer, founded by University of Washington professor Sreeram Kannan, became the sector's dominant platform. By August 2025, it held $22.06 billion in restaked assets and supported an ecosystem of 40 live AVSs with 190 more in development. The protocol attracted $171 million in venture funding, according to Tracxn data.
The capital inflows were driven substantially by points programs — pre-token incentive schemes that promised future airdrops to depositors. When EigenLayer's EIGEN token launched in late 2024, the points-to-token conversion triggered a withdrawal cycle. Slashing — the mechanism allowing AVSs to penalize misbehaving operators — went live on mainnet on April 17, 2025, introducing real financial risk to depositors for the first time without a corresponding increase in yield to compensate.
The gap between restaking's capital base and its revenue generation has widened to a level that no longer supports the sector's operating model.
| Metric | Restaking Sector | Liquid Staking Sector | |---|---|---| | TVL (Sept 8, 2026) | $10.02 billion | $51.87 billion | | Weekly Fees | $99,977 | $27.35 million | | Revenue per $1 secured (weekly) | $0.0000099 | $0.000527 | | Fee multiple vs. restaking | 1x | ~53x |
Ethereum base staking yield sat at approximately 2.5% annually as of September 23, 2026, according to Upshift Finance data. The additional yield from restaking — the core value proposition of the entire sector — has consistently failed to exceed a few hundredths of a percentage point above the base staking rate on a per-dollar basis.
On Kelp's financial statements, EIGEN token rewards appeared at $460,600 in gross revenue matched by $460,600 in cost of revenue: tokens arrived from EigenLayer and passed directly to depositors, leaving zero margin for the protocol, according to CoinDesk reporting.
Ether.fi launched in 2024 with a single product: it accepted ETH deposits and automatically restaked them on EigenLayer. Every dollar that entered the protocol was restaked by default. That model no longer exists.
In August 2026, ether.fi removed restaking from its flagship weETH token, converting it into an ordinary liquid staking token. Users who still want exposure to restaking must opt into a separate token built on Symbiotic, EigenLayer's smaller competitor. By September, less than 1% of ether.fi's assets remained restaked on EigenLayer. The company plans to remove its EigenPod withdrawal credentials entirely by the end of Q4 2026.
The revenue shift is measurable. Ether.fi's card fees — from its neobanking product — rose from 17% of monthly revenue in January 2026 to 46% in July. Restaking revenue, at $2.87 million, remained the second-largest profit line at the time of exit, but CEO Silagadze stated: "Neobank revenue has fully replaced the revenue lost from restaking and lower ETH price."
Ether.fi's gross profit declined 47%, from $18.71 million in Q3 2025 to $9.99 million in Q2 2026. The company claims a 38% annual revenue growth rate, but that growth is coming from financial services, not restaking.
On April 18, 2026, an attacker sent a crafted message to Kelp DAO's LayerZero-powered cross-chain bridge. The bridge accepted the message as legitimate and released 116,500 rsETH — approximately $292 million and roughly 18% of the token's circulating supply. No ETH changed hands on the other side; the rsETH was minted from nothing.
The attacker deposited the unbacked rsETH into Aave V3 and V4 as collateral and borrowed approximately $236 million in WETH against it, according to CoinDesk. Within 46 minutes, $5.4 billion of ETH and WETH had been withdrawn from Aave, WETH pool utilization hit 100%, and depositors were temporarily unable to withdraw. On-chain investigators linked the attack to North Korea's Lazarus Group, according to Chainalysis.
The Kelp exploit demonstrated a structural risk that restaking's design introduced: layered collateral creates layered contagion paths. A single bridge vulnerability in one LRT protocol cascaded into the largest lending platform in DeFi. Aave faced up to $230 million in potential bad debt from the incident, according to CoinDesk reporting.
The exploit was the largest DeFi hack of 2026 and served as a concrete demonstration of the cascading risks that critics had warned restaking would introduce since 2023.
The five largest remaining LRT protocols — those that continue to operate in the restaking business — generated the following combined results in Q2 2026:
| Protocol | Q2 2026 Gross Profit | Notable Context | |---|---|---| | Renzo | Not individually disclosed | Part of combined $953K figure | | Kelp | Not individually disclosed | Recovering from $292M exploit | | Swell | $22,370 | — | | Puffer Finance | $21,590 | Raised $23 million in venture funding | | Bedrock | Not individually disclosed | Part of combined $953K figure | | Combined | $953,350 | Down from $2.18M three quarters prior |
Puffer Finance's quarterly profit of $21,590 against $23 million in venture capital raised illustrates the gap between investment thesis and operating reality. At that profit rate, Puffer would need over 266 years to return its investors' capital, excluding operating expenses, team compensation, and infrastructure costs.
The profits that these protocols did generate came from ordinary staking fees charged underneath the restaking layer — not from restaking itself. The restaking component was, in accounting terms, a pass-through with zero margin.
In June 2025, Eigen Labs rebranded EigenLayer to EigenCloud and repositioned from a pure restaking protocol to a "verifiable cloud" platform. The new product suite includes EigenDA (data availability), EigenCompute (verifiable off-chain compute), EigenAI (AI inference), and EigenVerify (dispute resolution).
The rebrand represents an acknowledgment that restaking alone could not sustain the protocol's ambitions or its valuation. Sreeram Kannan told The Big Whale: "Our ambition goes far beyond restaking."
The EIGEN token tells a different story. From a 2024 peak of $5.65, EIGEN fell to a historical low of $0.1483 in April 2026 before recovering to approximately $0.26 as of late September. The 93.6% decline from peak to April's low represents one of the steepest drops among venture-backed Ethereum infrastructure tokens.
EigenLayer's TVL trajectory:
Symbiotic, the primary restaking competitor, holds approximately $329 million in TVL with 50+ integrated networks, 78 operators, and 55 vaults. The total restaking sector TVL, including all platforms, stands at approximately $10 billion — down from a collective peak above $23 billion.
The restaking narrative rested on a specific economic hypothesis: that AVS demand for shared security would generate fees sufficient to justify the additional risk and complexity of restaking over ordinary staking. Eighteen months after mainnet deployment, the data does not support that hypothesis.
Weekly fee generation of $99,977 across $10.02 billion in restaked capital means the sector earns approximately $5.2 million annually. Against that revenue, protocols must fund smart contract audits, monitoring infrastructure, bridge security, multisig coordination, and incident response teams. The Kelp exploit alone exceeded 56 years of the sector's total fee output.
The sector's trajectory follows a pattern consistent with subsidy-driven growth documented in the foundational economic value analysis of blockchain ecosystems. Capital entered during the points-farming phase, inflating TVL metrics that attracted more capital in a reflexive loop. When subsidies ended and slashing introduced real risk, the loop reversed.
EigenCloud's pivot to verifiable computing suggests the founding team recognizes the same conclusion the fee data implies: restaking-as-a-service does not generate enough revenue to sustain a standalone business. Whether verifiable computing — AI inference, off-chain compute, dispute resolution — can generate that revenue remains unproven. The market, as measured by the EIGEN token's 93.6% decline, has priced in significant doubt.
The restaking sector's contraction is not a temporary market downturn. It reflects a structural mismatch between the capital allocated to restaking and the fees that demand-side users — AVSs requiring shared security — are willing to pay. The sector attracted $22 billion in deposits during a points-driven cycle, but the underlying economic activity supports a fraction of that capital base.
Ether.fi's exit is the most visible signal, but it is not the only one. Protocol-level profitability data, token price performance, and the EigenCloud rebrand all point in the same direction: restaking, as originally conceived, has not produced the economic value its proponents projected. The technology may find a sustainable role within a broader verifiable computing stack. But the standalone restaking business — where additional yield was the product — has been repriced to near zero by the market.