Restaking & EigenLayer Risks

What Is Restaking?

After Ethereum's transition to Proof of Stake in September 2022, validators began locking 32 ETH to participate in consensus. That staked ETH earns a yield of approximately 3-4% annually for performing validation duties: attesting to blocks, proposing blocks, and participating in sync committees. The capital is productive, but it serves only one purpose — securing Ethereum's base layer.

Restaking takes that already-staked ETH and extends its security guarantees to additional protocols. Instead of your 32 ETH only securing Ethereum consensus, it can simultaneously secure a data availability layer, a cross-chain bridge, an oracle network, or a rollup sequencer. These additional protocols are called Actively Validated Services (AVS).

The core premise is capital efficiency. Rather than each new protocol bootstrapping its own independent validator set with its own staked token (which is expensive and results in fragmented security), protocols can borrow Ethereum's existing economic security. Validators opt in to validate additional services and, in exchange, earn supplemental yield on top of their base Ethereum staking rewards.

EigenLayer, launched on Ethereum mainnet in mid-2023, is the dominant restaking platform. It introduced the concept and has attracted tens of billions of dollars in total value locked at its peak. Other restaking platforms have emerged — Symbiotic and Karak among them — but EigenLayer remains the market leader and the primary source of both opportunity and risk in the restaking ecosystem.

Why This Matters

Restaking is not simply another DeFi primitive. It fundamentally changes the relationship between Ethereum's security and the broader ecosystem. If restaking succeeds at scale, Ethereum becomes the trust layer for dozens or hundreds of protocols. If it fails — through a major slashing event or correlated collapse — it could damage Ethereum's own security guarantees. The stakes, both literally and figuratively, are higher than most DeFi innovations.

How EigenLayer Works

EigenLayer functions as a middleware layer between Ethereum validators and AVS protocols. At a high level, the system works through three coordinated components: operators, delegators, and AVS modules.

Operators

Operators are entities that run the actual validation software for AVS protocols. They register with EigenLayer, declare which AVS they will validate, and run the necessary node software. Operators can be the same entities running Ethereum validators, or they can be specialized restaking operators. Each operator must meet the technical requirements of every AVS they opt into — running specific software, maintaining uptime, and executing the validation logic defined by that AVS.

Delegators (Restakers)

Delegators are ETH holders who stake their ETH (or Liquid Staking Tokens like stETH) into EigenLayer and delegate it to operators. They earn yield from the AVS protocols their chosen operator validates, but they also share in the slashing risk. Delegators trust operators to behave honestly and run their infrastructure correctly. If an operator misbehaves on an AVS, the delegated stake is subject to slashing.

Actively Validated Services (AVS)

AVS are the protocols that borrow Ethereum's security through EigenLayer. Each AVS defines its own validation logic, performance requirements, and slashing conditions. Examples of AVS categories include:

• Data availability layers: EigenDA is EigenLayer's own data availability service, providing low-cost data availability for rollups.
• Cross-chain bridges: Bridges can use restaked ETH to secure message passing between chains, replacing or supplementing their own validator sets.
• Oracle networks: Price feed and data oracle protocols can leverage restaked security instead of maintaining independent staking mechanisms.
• Rollup sequencers: Decentralized sequencing for Layer 2 rollups, where restaked validators take turns proposing transaction orderings.
• Keeper networks: Automated task execution services (liquidations, rebalancing) secured by restaked collateral.

Slashing Mechanics

The critical difference between staking and restaking is the slashing surface. In regular Ethereum staking, your ETH can only be slashed for Ethereum-specific violations: double-signing a block or making contradictory attestations. These conditions are well-defined, rare, and well-understood after years of operation.

With restaking, each AVS adds its own slashing conditions. If a validator signs an incorrect data availability attestation for EigenDA, or provides a fraudulent price feed for an oracle AVS, or fails to meet uptime requirements for a sequencer AVS, the restaked ETH can be slashed. The same ETH is now exposed to slashing risk from Ethereum consensus and every AVS the operator has opted into.

The Risk Stack

Restaking creates a layered risk structure where each layer compounds the risks of the layers beneath it. Understanding this stack is critical for evaluating whether the additional yield justifies the exposure.

Layer 1: Ethereum Staking Risk

The base layer is standard Ethereum staking risk. This includes protocol-level slashing for consensus violations, inactivity penalties for extended offline periods, and smart contract risk in the deposit contract. After several years of operation, these risks are well-characterized. Slashing events on Ethereum mainnet have been rare — measured in dozens of incidents across hundreds of thousands of validators. The penalty structure is understood, validator client diversity is improving, and the economic incentives are battle-tested. This layer is the foundation, and it is relatively solid.

Layer 2: AVS-Specific Risk

Each AVS introduces its own risk profile. This includes smart contract bugs in the AVS's slashing logic (what if the contract incorrectly identifies honest behavior as malicious?), misconfigured slashing parameters (thresholds set too aggressively), and operational complexity (running multiple validator clients simultaneously increases the chance of misconfiguration). Because most AVS protocols are new and have not been stress-tested through market crises, their slashing conditions are less proven than Ethereum's. A bug that triggers incorrect slashing on an AVS could affect all restakers who opted into that service.

Layer 3: Correlated Risk

This is the most concerning layer. If multiple AVS experience issues simultaneously — perhaps because they share a common software dependency, or because a systemic event (like an Ethereum reorganization) triggers slashing across multiple services at once — the result is mass slashing. Large-scale correlated slashing would remove significant ETH from Ethereum's security budget, potentially weakening the chain's consensus security at precisely the moment when it is most needed.

Consider a scenario: a critical bug is discovered in a widely-used AVS client library that causes validators to sign incorrect attestations. If 30% of restaked ETH is validating services that use this library, a correlated slashing event could remove billions of dollars of security from Ethereum's validator set. The resulting validator exits and market panic could trigger a cascading effect far beyond the AVS ecosystem.

The Security Budget Math

Restaking’s economic security model can be understood through two key metrics:

• Cost of Corruption (CoC) — The minimum value of stake that must be slashed to compromise an AVS. This equals the sum of all colluding operators’ stakes multiplied by their slashing fractions.
• Profit from Corruption (PfC) — The maximum extractable value from compromising the AVS (e.g., draining a bridge, manipulating an oracle).

Security requires that CoC > PfC — it must be economically irrational to attack. This is the fundamental invariant that the entire restaking model depends on.

ETH Staking Landscape (February 2026)

The scale of restaking risk must be understood in context. Total staked ETH: ~34.4 million ETH (~28% of supply). The distribution reveals concentration risks:

EtherFi is now the largest liquid restaking protocol, with 2.15M ETH making it the third-largest staking entity overall. This concentration means restaking risk is not evenly distributed — a significant portion of Ethereum’s security budget flows through a handful of protocols.

Babylon: BTC Restaking

Restaking is expanding beyond Ethereum. Babylon Protocol enables native Bitcoin staking using a cryptographic technique called Extractable One-Time Signatures (EOTS) — no bridging, no wrapping, no third-party custody. BTC remains on the Bitcoin network while providing economic security to Cosmos-based PoS chains. This represents the first mechanism for Bitcoin holders to earn yield on their BTC without giving up custody or moving it to another chain.

Liquid Restaking Tokens (LRTs)

Just as Liquid Staking Tokens (like Lido's stETH) made staked ETH liquid and composable in DeFi, Liquid Restaking Tokens do the same for restaked ETH. LRT protocols accept ETH or LST deposits, restake them through EigenLayer, and issue a liquid receipt token that represents the restaked position.

Major LRT Protocols

• EtherFi (eETH): The largest LRT protocol by TVL. Users deposit ETH and receive eETH, which can be used across DeFi while the underlying ETH earns both Ethereum staking yield and EigenLayer restaking yield. EtherFi was the first LRT to launch its own governance token.
• Renzo (ezETH): A cross-chain restaking protocol that deploys restaked ETH across multiple AVS strategies. Renzo's ezETH token experienced a notable depeg event in April 2024 when it briefly traded at a discount to ETH, highlighting the liquidity risks inherent in LRTs.
• Kelp (rsETH): Manages restaking strategies and issues rsETH as a receipt token. Kelp focuses on optimizing AVS selection to balance yield and risk across its restaking portfolio.
• Puffer Finance (pufETH): Combines native restaking with anti-slashing technology, using secure hardware (Intel SGX) to reduce the risk of slashing events for its validators.

The Additional Smart Contract Layer

LRTs add yet another layer of smart contract risk on top of the existing restaking stack. The risk tower now includes: the Ethereum deposit contract, the EigenLayer core contracts, the AVS-specific contracts, and the LRT protocol's own contracts. Each contract layer is a potential point of failure. An exploit in an LRT protocol's withdrawal logic, for instance, could lock user funds regardless of whether the underlying restaking is functioning correctly.

Depeg Risk and DeFi Composability

LRT tokens are widely used as collateral in DeFi lending protocols, as liquidity in DEX pools, and in leveraged yield strategies. This composability creates a specific danger: if an LRT depegs from its expected value — whether due to a slashing event, a liquidity crisis, or a smart contract bug — the consequences cascade through every DeFi protocol that holds it as collateral. Lending protocols face bad debt. Leveraged positions get liquidated. DEX pools become imbalanced. The 2024 ezETH depeg, while brief, demonstrated how quickly LRT price dislocations can propagate through DeFi.

Governance & Centralization Concerns

EigenLayer occupies a uniquely powerful position in the Ethereum ecosystem. It controls the middleware through which a large and growing share of staked ETH is redirected to secure external protocols. This concentration of control raises several governance questions.

Upgrade Authority

EigenLayer's smart contracts are upgradeable, and the protocol's multisig and advisory committees have the authority to modify contract parameters, adjust slashing conditions, and implement protocol changes. While upgradeability is necessary for a complex and evolving protocol, it means that a small group of individuals can make changes that affect billions of dollars of restaked ETH. There is no on-chain governance vote required for many of these changes. The community must trust the EigenLayer team to exercise this authority responsibly.

Token Distribution Controversies

The EIGEN token launch in 2024 generated significant controversy. Initial distribution was restricted by geography (excluding US users from claiming), the token was initially non-transferable, and the airdrop allocation was perceived by many participants as insufficient relative to the capital they had committed. These governance decisions affected user trust and highlighted the tension between a protocol that borrows Ethereum's decentralized security while operating with centralized decision-making about its own token economics.

Systemic Importance

As EigenLayer grows, it becomes systemically important to Ethereum itself. If a significant percentage of all staked ETH is restaked through EigenLayer, then EigenLayer's operational decisions — which AVS to support, how to handle slashing disputes, whether to pause contracts in an emergency — have consequences for Ethereum's security. This creates a single point of influence that sits uncomfortably with Ethereum's ethos of decentralization and credible neutrality. Whether one private company should control the security delegation layer for the world's largest proof-of-stake network is an open and unresolved question.

The Bull Case

Proponents of restaking argue that it could be the most significant economic innovation for Ethereum since the merge to Proof of Stake. The bull case rests on several interconnected arguments.

Shared Security Reduces Bootstrapping Costs

Launching a new protocol that requires economic security is expensive. Without restaking, a new oracle network or bridge must issue its own token, incentivize validators to stake it, and hope that enough capital is committed to make attacks economically infeasible. This bootstrapping problem is a major barrier to entry. Restaking solves it by allowing new protocols to rent Ethereum's established security from day one. A new AVS can launch with billions of dollars of security backing it immediately, without needing to build a validator set from scratch.

Higher Yield Attracts More Staking

If restaking reliably generates additional yield beyond base Ethereum staking, it creates a positive feedback loop. Higher yields attract more ETH to be staked, which increases Ethereum's total security budget, which makes the network more secure, which makes it more valuable, which attracts more staking. This flywheel effect could significantly increase the percentage of ETH supply that is staked, from roughly 27% today to potentially 40% or more.

Ethereum as the Universal Security Layer

The most ambitious bull case positions Ethereum not just as a smart contract platform, but as the security infrastructure for the entire crypto ecosystem. If every bridge, oracle, data availability layer, and rollup sequencer can borrow Ethereum's security through restaking, then Ethereum becomes indispensable — the cryptographic trust layer that everything else relies on. This would create enormous demand for ETH as a staking asset and could justify significantly higher valuations for the network.

Capital Efficiency

From a capital allocation perspective, restaking is undeniably efficient. The same 32 ETH can secure Ethereum consensus, provide data availability, validate oracle feeds, and sequence rollup transactions — all simultaneously. Without restaking, each of these functions requires its own independent capital lockup. The reduction in total capital required to secure the same set of services is a genuine economic benefit.

The Bear Case

Critics of restaking raise concerns that range from technical risk to philosophical objections about the nature of Ethereum's security model.

Leverage on Ethereum's Security

Restaking is fundamentally a form of leverage. The same ETH is being used to secure multiple systems simultaneously. In good times, this leverage creates efficiency. In bad times, it amplifies losses. If a major slashing event occurs across multiple AVS, the same ETH gets slashed multiple times, potentially removing a disproportionate amount of security from every system it was backing — including Ethereum itself. This is not unlike how fractional reserve banking creates efficiency in normal times but systemic fragility during crises.

Correlated Risk Threatens Ethereum

If restaking losses are correlated — and the shared infrastructure, shared operators, and shared software dependencies suggest they could be — then Ethereum's own security is at risk. A cascading slashing event could force large numbers of validators to exit the network, reducing the cost of attacking Ethereum's consensus. The network that was supposed to provide security to the ecosystem could itself become less secure because of restaking.

Complexity Without Proportional Benefit

The additional yield from restaking is currently modest — often just 1-3% above base staking yield from actual AVS payments (excluding point speculation). Critics argue that this incremental yield does not justify the exponential increase in complexity and risk. The smart contract surface area, the operational burden of running multiple validator clients, and the governance centralization all represent costs that may outweigh the benefits for most participants.

The "Hotel California" Problem

During market stress, everyone wants to exit at the same time. Restaking withdrawal queues, EigenLayer's unbonding period, and potential LRT depeg events can all create situations where users cannot exit their positions quickly. If a slashing event is detected, the rational response is to withdraw immediately — but if everyone attempts this simultaneously, the withdrawal queue extends, LRT liquidity evaporates, and users are trapped in a deteriorating position. The ability to enter a restaking position easily does not guarantee the ability to exit one.

Key Takeaways