EigenLayer and Ethereum Restaking: Yield Opportunities vs Systemic Risk

Ethereum restaking lets stakers pledge their already-staked ETH to secure additional protocols, earning extra yield on the same capital. EigenLayer, the protocol that pioneered this model, grew from under $1 billion in total value locked to over $15 billion in early 2025. It then lost more than half that TVL within months of launching its slashing mechanism. The trajectory illustrates a central tension: restaking creates real yield opportunities, but it also introduces systemic risks that Ethereum has never faced before.

This article explains how EigenLayer works at the contract level, where restaking yield actually comes from, what happens when leverage becomes recursive, and how Bitcoin restaking via Babylon takes a fundamentally different approach to the same problem.

How EigenLayer Restaking Works

EigenLayer extends Ethereum's proof-of-stake security model by allowing stakers to opt into securing additional services called Actively Validated Services (AVSs). In exchange for taking on additional slashing risk, restakers earn rewards from each AVS they help secure. The core insight: Ethereum validators already have staked capital at risk, so that same capital can simultaneously back other protocols, provided the staker accepts the additional slashing conditions.

Smart Contract Architecture

The protocol consists of several interlocking contracts deployed on Ethereum mainnet:

• DelegationManager: the central orchestration contract handling operator registration, staker delegation, undelegation, and withdrawal processing
• StrategyManager: manages deposits of liquid staking tokens (stETH, rETH, cbETH) through ERC-4626 vault strategies, with each LST having a dedicated strategy contract
• EigenPodManager and EigenPod: facilitates native beacon chain ETH restaking, where each staker gets an individual EigenPod contract that serves as their validator's withdrawal credentials
• AllocationManager: manages Operator Sets and handles operator-to-AVS registration, replacing the earlier AVSDirectory
• PaymentCoordinator: handles AVS-to-participant reward distribution through Merkle tree-based claim verification

Two Restaking Methods

Stakers can restake through two paths. Native restaking requires running a beacon chain validator and pointing its withdrawal credentials to an EigenPod contract. This requires the full 32 ETH minimum but is the most capital-efficient approach. LST restaking deposits liquid staking tokens like stETH or rETH into the StrategyManager. It is more accessible but adds a layer of smart contract dependency.

In both cases, stakers delegate their shares to an operator through the DelegationManager. Operators then register with specific AVSs through Operator Sets, each with defined responsibilities and slashing conditions. Withdrawals carry a mandatory 14-day waiting period to ensure irregular behavior can be penalized before funds are released.

The AVS Ecosystem

Actively Validated Services are the protocols that consume restaked security. By February 2026, EigenLayer had 156 registered AVSs with approximately 40 live on mainnet and over 190 in development. The protocol rebranded AVS from "Actively Validated Services" to "Autonomous Verifiable Services" as part of a broader pivot toward what it calls EigenCloud.

A growing category of AI verification AVSs has emerged, with projects seeking trust-minimized model evaluation secured by restaked ETH. The trend toward specialized "Vertical AVSs" reflects a market where generic security pooling is giving way to purpose-built validation services.

Where Restaking Yield Comes From

Understanding the sustainability of restaking requires separating real yield from speculative incentives. Restaking yield has four sources, each with different durability:

1. Ethereum consensus rewards: the base staking yield of approximately 3-4% APY from validating the beacon chain
2. AVS protocol fees: payments from AVSs to operators for security services, adding an estimated 1-3% annualized yield depending on which AVSs are selected
3. EIGEN token emissions: programmatic incentive rewards that supplement real fee revenue, historically the largest component of restaking "yield"
4. Operator commissions: operators typically charge 5-8% on rewards passed through to stakers

For passive restakers, total returns typically range from 3.8% to 6% APY. With DeFi leverage strategies involving recursive borrowing loops on platforms like Aave or Morpho, effective APY can reach 12-20%, but with significant liquidation risk.

The sustainability question: AVSs would need less than 10% of EigenLayer's TVL for their actual security needs. This means most of the capital in the protocol is not generating organic yield. EigenLayer's protocol revenue fell from $31.93 million in Q1 2025 to $8.74 million in Q1 2026: a 73% decline that highlights the gap between speculative capital inflows and real demand for restaked security.

The Liquid Restaking Token Ecosystem

Just as liquid staking tokens like stETH made staked ETH composable in DeFi, liquid restaking tokens (LRTs) do the same for restaked ETH. Protocols like EtherFi, Renzo, Puffer, Kelp, and Swell issue tokens representing restaked positions, allowing holders to use them as collateral elsewhere while earning restaking yield.

By early 2026, the total LRT market held approximately 3.34 million ETH. EtherFi's weETH earned the first independent A+ risk rating in liquid restaking. But this market is not without incident: in April 2024, Renzo's ezETH depegged by 18.3% on Uniswap following controversial tokenomics changes, triggering cascading liquidations across Gearbox and Morpho Blue. The token re-pegged within hours, but the episode demonstrated exactly the kind of correlated failure that critics had warned about.

Systemic Risks: The Case Against Recursive Leverage

Restaking introduces risks that go beyond individual protocol failure. The concern is not just that one AVS might misbehave, but that the layered structure of restaking creates amplification effects that could propagate through the broader Ethereum ecosystem.

Recursive Leverage

The most discussed systemic risk is recursive leverage. A staker deposits ETH into a validator, receives stETH, deposits that into EigenLayer, receives an LRT like eETH, deposits that LRT as collateral on Aave or Morpho, borrows ETH, and restakes again. Each layer amplifies the effective yield but also amplifies exposure to any disruption in the chain.

EigenLayer itself disallows depositing LRTs back into the protocol. But DeFi markets operate independently, and leveraged LRT positions are common on lending platforms. The result is a tower of claims all resting on the same underlying ETH. If any layer experiences a disruption (an LRT depeg, a slashing event, a smart contract bug) the forced unwinding cascades through every layer above it.

Correlated Slashing

Slashing went live on EigenLayer's mainnet on April 17, 2025. Each AVS independently determines its slashing conditions, and operators opt into those conditions through Operator Sets. The AllocationManager introduced "Unique Stake" allocation, where operators can earmark portions of delegated stake for specific AVSs so that only one AVS can slash that particular portion.

The concern is correlation. A bug in a widely adopted AVS could trigger mass slashing simultaneously across many operators. If those operators are also securing other AVSs, the capital reduction from one slashing event weakens security guarantees elsewhere. If the slashed assets were also serving as collateral on lending platforms, the liquidation cascade compounds the damage.

The "Too Big to Slash" Problem

EigenLayer commands approximately 75-94% of the Ethereum restaking market depending on measurement methodology. With over 1,200 operators and billions in restaked ETH, a large-scale slashing event would not just affect restakers: it could destabilize DeFi protocols that accept LRTs as collateral, bridges that rely on restaked security, and potentially even Ethereum's own consensus layer if validators lose significant stake.

In May 2023, Vitalik Buterin addressed this risk directly in a blog post titled "Don't overload Ethereum's consensus". He warned that using Ethereum's consensus layer for restaking purposes "could bring high systemic risks to the ecosystem and should be discouraged and resisted." His argument: dual-use of validator staked ETH is "fundamentally fine," but expanding validator duties increases costs, complexities, and risks to the base layer.

Academic research confirms the concern: A 2026 IEEE paper titled "Systemic Risk in Ethereum's Restaking Architecture" found that "the reuse of collateral combined with centralized delegation and complex smart contract architectures creates conditions for risk amplification and correlated failures." The study also found that many AVSs are secured by highly centralized validator sets, undermining the promise of pooled security.

Conflict of Interest Controversy

The systemic risk debate took on a governance dimension in 2024 when Ethereum Foundation researchers Justin Drake and Dankrad Feist accepted advisory roles with EigenLayer, reportedly with significant token compensation. The conflict of interest was clear: researchers shaping Ethereum's roadmap were financially tied to a protocol whose growth depended on that roadmap's direction. Both resigned from their EigenLayer advisory roles in November 2024 following community backlash.

EigenLayer's Market Trajectory

The protocol's trajectory since slashing went live tells its own story. TVL peaked above $15 billion in early 2025, then fell to approximately $7 billion by late 2025 after slashing introduced real economic risk. Protocol revenue declined 73% year-over-year. The EIGEN token, which reached an all-time high of $5.65 in December 2024, traded around $0.24 by mid-2026.

Eigen Labs responded by rebranding to EigenCloud in June 2025, positioning the platform as a "verifiable cloud" encompassing data availability (EigenDA), compute, AI inference, and dispute resolution. The company secured a $70 million token purchase from a16z to support the pivot. A month later, Eigen Labs cut 25% of its workforce (29 employees) to refocus resources. Under a new incentive framework (ELIP-12), only fee-paying AVSs qualify for incentive rewards, and a portion of fees is directed toward EIGEN token buybacks.

Competition has also emerged: Symbiotic launched on mainnet in January 2025 with slashing enabled from day one, reaching approximately $897 million in TVL and capturing around 5.5% of the restaking market.

Bitcoin Restaking: A Different Architecture

Babylon Protocol applies the restaking concept to Bitcoin, but with an architecture that avoids many of the systemic risks present in Ethereum's model. As of mid-2026, Babylon holds approximately 56,800 BTC (around $5.6 billion), making it the largest Bitcoin staking protocol by TVL.

How Babylon Avoids Recursive Leverage

The key architectural difference: Babylon keeps BTC natively on the Bitcoin blockchain. There is no bridging, no wrapping into ERC-20 tokens, and no third-party custody. Staked BTC remains in a self-custodial UTXO with two spending conditions: a timelock after which the staker can withdraw using their secret key, and a burning path through Extractable One-Time Signatures (EOTS).

Because the staked BTC stays as a Bitcoin UTXO rather than becoming a composable Ethereum-side token, it cannot easily be deposited as collateral on Aave, Morpho, or other DeFi lending platforms. This eliminates the recursive leverage loop at the architectural level. You cannot build a tower of claims on top of Babylon-staked BTC the way you can with EigenLayer LRTs.

The caveat: Liquid staking derivatives built on top of Babylon, such as Lombard's LBTC, could reintroduce similar leverage risks. In April 2025, Lombard temporarily unstaked 14,929 BTC during a finality provider transition, causing a 32% TVL drop. The recursive leverage problem may follow wherever liquid derivative tokens exist, regardless of the underlying chain.

EOTS: Slashing Without Smart Contracts

Babylon uses Extractable One-Time Signatures for slashing, a cryptographic mechanism built on Schnorr signatures natively supported by Bitcoin via Taproot. An EOTS key can safely sign one message, but if used twice for conflicting messages, anyone can mathematically extract the private key. If a validator double-signs (a safety violation), the revealed key allows anyone to broadcast a slashing transaction that burns the staked BTC.

This design achieves enforceable slashing on Bitcoin without smart contracts: something that would have seemed impossible before Taproot enabled Schnorr signature support. A block is finalized only after receiving EOTS signatures from more than two-thirds of the staked BTC weight.

EigenLayer vs Babylon: Restaking Models Compared

The architectural distinction matters. EigenLayer's power comes from composability: the same restaked ETH can simultaneously secure dozens of AVSs while the LRT representing it circulates through DeFi. Babylon sacrifices that composability for isolation. A Bitcoin staker in Babylon earns yield from the BSNs they help secure, but the staked capital cannot be rehypothecated across DeFi lending markets. This limits yield potential but also limits the blast radius of any failure.

What This Means for Bitcoin Layer 2s

The Ethereum restaking experience offers important lessons for the Bitcoin ecosystem. Bitcoin's lack of native smart contract composability, often cited as a limitation, turns out to be a structural advantage when it comes to avoiding recursive leverage. Protocols building on Bitcoin, from Babylon to Spark, inherit this property by default.

Spark's approach to Bitcoin scaling avoids restaking entirely. Rather than reusing staked capital to bootstrap security for additional services, Spark uses statechains and FROST threshold signatures to enable instant, self-custodial transfers without the layered economic dependencies that characterize Ethereum's restaking stack. There are no liquid derivative tokens, no recursive collateral loops, and no cascading slashing risk. The tradeoff is different: Spark relies on operator honesty (1-of-n security) rather than economic penalties, but the failure mode is unavailability rather than systemic contagion.

For developers building on Bitcoin, the Spark SDK and documentation provide a path to building payment applications without exposure to restaking complexity. The Bitcoin restaking landscape continues to evolve, but the design choices being made today will determine whether Bitcoin's Layer 2 ecosystem avoids the recursive leverage patterns that have strained Ethereum's.

Key Takeaways

Restaking is not inherently dangerous. The concept of reusing staked capital to secure additional services is sound in principle. The risk emerges from the interaction effects: liquid derivative tokens that enable recursive leverage, concentrated operator sets that create correlated slashing exposure, and incentive structures that attract speculative capital far beyond what AVSs actually need for security.

EigenLayer's evolution from a $15 billion protocol to one navigating a rebrand, workforce reduction, and 73% revenue decline is not a failure story. It is a stress test of a novel economic mechanism. The Unique Stake allocation system, fee-based incentive reforms, and multi-chain expansion all represent attempts to build sustainable infrastructure on top of the restaking primitive.

The question is whether the yield from AVS fees will ever justify the capital deployed, or whether restaking will remain dependent on token emissions and speculative positioning. For Ethereum, the answer shapes the risk profile of the entire DeFi ecosystem. For Bitcoin, the answer informs which architectural patterns to adopt and which to avoid.

This article is for educational purposes only. It does not constitute financial or investment advice. Bitcoin and Layer 2 protocols involve technical and financial risk. Always do your own research and understand the tradeoffs before using any protocol.