Shared security is a misnomer. The term implies a unified, fungible security layer, but EigenLayer and its AVS ecosystem create isolated security silos. Each Actively Validated Service (AVS) negotiates its own slashing conditions and economic guarantees with operators.
liquid-staking-and-the-restaking-revolution
Blog
Why Shared Security is a Misnomer in the Restaking Revolution
The term 'shared security' falsely implies a common pool. In reality, restaking creates competitive markets where capital is explicitly re-allocated, not shared. This analysis breaks down the economic and security implications for architects.
introduction
THE MISNOMER
Introduction: The Shared Security Lie
The term 'shared security' is a marketing misnomer that obscures the fundamental economic and technical fragmentation of the restaking ecosystem.
The security is not shared, it's rented. Node operators allocate specific capital to specific AVSs, creating a fragmented marketplace. The security of an EigenDA data availability layer is not the same as the security of an Omni Network cross-chain messaging AVS.
This fragmentation creates systemic risk. A correlated slashing event in one AVS does not automatically cascade, but it triggers mass operator exits that drain liquidity from all interconnected services. The security model is interdependent, not shared.
Evidence: The rapid proliferation of AVSs—from AltLayer rollups to Lagrange coprocessors—proves the demand for cryptoeconomic security. However, each one must bootstrap its own bespoke operator set and slashing contract, replicating, not sharing, the security bootstrap problem.
key-insights
RESTAKING REALITIES
Executive Summary: Three Hard Truths
The narrative of 'shared security' in restaking is a marketing illusion that obscures systemic risks and capital inefficiency.
01
The Problem: Security is Not a Liquid Commodity
Security is a non-fungible property derived from validator slashing and social consensus. EigenLayer's restaking model treats it as a transferable resource, creating a dangerous abstraction.\n- Security is contextual: A validator's stake securing Ethereum is not the same asset securing an Omni Network AVS.\n- Dilution risk: $20B+ in restaked ETH does not mean $20B of security for each service; it's shared, divisible, and contingent.
1 Asset
N Fungible Claims
$20B+ TVL
Shared Pool
02
The Solution: Verifiable Security Budgets
Protocols must audit and quantify the actual security budget allocated to them, not the total restaking pool. This shifts the focus from TVL to cryptoeconomic guarantees.\n- Explicit staking: Projects like Babylon are pioneering Bitcoin timestamping and staking for PoS chains, creating direct, verifiable security.\n- Risk pricing: EigenLayer's upcoming slashing marketplace is an attempt to price risk, but it's a reactive, not preventative, mechanism.
Auditable
Security Budget
Reactive
Slashing Market
03
The Reality: Systemic Correlation Trumps Isolation
The 'isolation' of slashing faults across Active Validation Services (AVSs) is a theoretical construct. In a crisis, social consensus and market panic create a single point of failure.\n- Contagion vector: A major slashing event on a high-profile AVS (e.g., EigenDA, Lagrange) could trigger mass unstaking, collapsing security for all.\n- Layer 0 Risk: This creates a new systemic risk layer atop Ethereum, contradicting the modular 'reduce complexity' thesis.
1 Failure
Cascading Risk
New Layer 0
Systemic Risk
thesis-statement
THE MISNOMER
Core Thesis: Re-Allocation, Not Sharing
Restaking does not create new security; it re-allocates existing Ethereum staked capital to secure new protocols.
Shared security is a misnomer. The term implies a net increase in total security, but the capital securing EigenLayer AVSs is the same capital securing Ethereum. The protocol re-allocates slashing risk from Ethereum consensus to off-chain services like AltLayer or EigenDA.
Security is not shared, it is divided. A validator's stake cannot simultaneously provide full security to both Ethereum and an AVS. The economic commitment is split, creating a zero-sum re-allocation of slashable value rather than a multiplicative security boost.
The innovation is capital efficiency. Protocols like EigenLayer and Babylon unlock utility for idle stake. This is a superior model to bootstrapping new validator sets, as seen with Cosmos zones, but it trades off with systemic risk concentration on Ethereum.
Evidence: Ethereum's ~$100B staked ETH is the sole backstop. Every dollar secured for an EigenLayer rollup is a dollar of reduced economic security for Ethereum's base layer, a trade-off quantified by slashing condition design.
THE RESTAKING REALITY CHECK
Security Models: Shared Pool vs. Re-Allocation Market
Deconstructing the capital efficiency and risk isolation of two dominant security models in restaking, moving beyond the 'shared security' marketing.
| Core Feature / Metric | Shared Pool Model (EigenLayer) | Re-Allocation Market (Karak) | Direct Delegation (Native Staking) |
|---|---|---|---|
Primary Security Source | Rehypothecated ETH staking yield | Rehypothecated ETH staking yield | Native protocol token stake |
Capital Multiplier (Theoretical) |
|
| 1x (Direct stake) |
Operator Slashing Scope | Global pool slashing (Correlated risk) | Isolated AVS slashing (Compartmentalized risk) | Protocol-specific slashing |
Liquid Restaking Token (LRT) Role | Passive yield receipt (e.g., ezETH) | Active reallocation voucher (Capital routing) | N/A |
Yield Source for AVSs | Siphons from base staking yield | Direct bid from AVS treasury (Premium) | Native token emissions |
Liquidity Fragmentation Risk | High (LRT wars, peg instability) | Medium (Market-driven LRT valuation) | Low |
Time to Activate Security | ~7 days (EigenLayer queue + withdrawal) | < 1 day (Instant market allocation) | Varies by chain (e.g., 32 ETH activation) |
Exemplar Protocols | EigenLayer, Ether.fi, Renzo | Karak, Inception | Cosmos Hub, Solana, Ethereum |
deep-dive
THE REALITY
The Mechanics of Security Re-Allocation
Shared security is a marketing term; the restaking revolution is about the dynamic, permissionless re-allocation of cryptoeconomic security.
Security is a commodity that protocols like EigenLayer and Babylon unbundle from Ethereum consensus. Validators sell their staked ETH's slashing risk to new networks, creating a competitive marketplace for security.
The misnomer is 'shared'. Security is not pooled; it is auctioned and delegated. Each actively validated service (AVS) like AltLayer or EigenDA bids for a specific slice of a validator's stake, creating isolated slashing conditions.
This re-allocation is non-fungible. The security backing a data availability layer differs from an oracle network. The validator's capital is fragmented across unique risk profiles, not shared in a common pool.
Evidence: EigenLayer's mainnet holds over $15B in restaked ETH, with operators now choosing between dozens of AVSs. This capital flow defines a new security-as-a-service economy.
risk-analysis
WHY SHARED SECURITY IS A MISNOMER
Emerging Risks of the Re-Allocation Model
The promise of 'shared security' in restaking is a dangerous oversimplification that masks systemic fragility and concentrated risk.
01
The Problem: Correlated Slashing Cascades
Restaking pools like EigenLayer create a network of interdependent slashing conditions. A failure in one AVS (e.g., a data availability layer) can trigger a mass slashing event across hundreds of validators, propagating failure instead of containing it.
- Systemic Risk: A single bug or governance attack can cascade through the entire restaking ecosystem.
- Contagion Model: Unlike isolated chains, slashing is now a network contagion risk, not a local penalty.
100+
AVS Dependencies
>90%
Correlated Risk
02
The Problem: Liquidity Fragmentation & Capital Inefficiency
Capital is not 'shared'—it is fragmented and over-leveraged. The same staked ETH is simultaneously backing multiple services (rollups, oracles, bridges), creating a web of unsecured liabilities exceeding the underlying collateral.
- Capital Multiplier Illusion: TVL metrics are deceptive; the same ETH is counted multiple times across different AVSs.
- Liquidity Black Holes: A crisis triggers mass unbonding across all services simultaneously, freezing the entire ecosystem.
$10B+
Rehypothecated TVL
7-21 Days
Unbonding Lock
03
The Problem: Centralized Points of Failure
The 'shared' security layer consolidates power into a few operator cartels. Top node operators (e.g., Figment, Chorus One) run the majority of AVS nodes, creating a centralized failure layer that defeats decentralization goals.
- Operator Concentration: A handful of entities control the physical infrastructure for most restaked services.
- Single-Point Governance: EigenLayer's multisig and upgrade keys represent a centralized veto over the entire 'shared' security stack.
<10
Dominant Operators
8/11
Multisig Threshold
04
The Solution: Isolated Security Pools
Protocols like Babylon are pioneering Bitcoin timestamping and dedicated staking pools where security is opt-in and siloed. Failure in one pool does not bleed into others, preserving systemic integrity.
- Risk Containment: Slashing and faults are isolated to the specific pool and asset class.
- Clear Liability: Capital providers know exactly what chain or service they are securing, with no hidden interdependencies.
Zero
Cross-Pool Contagion
1:1
Collateral Ratio
05
The Solution: Explicit Risk Markets
Platforms need to evolve into risk quantification engines like Nexus Mutual or Sherlock, where AVSs purchase slashing insurance from dedicated capital pools. Security becomes a priced commodity, not a vague promise.
- Priced Risk: Operators and AVSs pay premiums based on their fault probability, creating a market signal for safety.
- Capital Specialization: Insurance providers can underwrite specific risk types (e.g., oracle failure vs. DA censorship).
Basis Points
Risk Premium
Actuarial
Pricing Model
06
The Solution: Modular Slashing & Verifiable Trust
Adopt a modular slashing framework where penalties are proportional, verifiable, and contestable. Projects like Obol's Distributed Validator Technology (DVT) and Lagrange's ZK proofs for state committees move fault detection on-chain.
- Proportional Penalties: Slashing is granular (e.g., 5% for a minor fault) instead of binary total loss.
- ZK-Proofed Faults: Use zero-knowledge proofs to verifiably prove a slashable offense, removing subjective governance.
ZK Proofs
Fault Verification
DVT
Base Layer
counter-argument
THE MISNOMER
Counterpoint: The Pooling of *Slashing* is Shared
Shared security in restaking pools is a marketing term that obscures the asymmetric concentration of slashing risk.
Risk is not uniformly distributed. The 'shared' in shared security refers to the aggregation of capital, not the equitable distribution of slashing penalties. A single operator's fault triggers a penalty that is proportionally applied across all stakers in that pool, creating a systemic contagion vector.
The slashing risk is pooled, not shared. This is a critical semantic distinction. Protocols like EigenLayer and Symbiotic create pools where slashing is a common-pool resource problem. The risk is mutualized, but the control and fault-tolerance remain with a few node operators, creating a principal-agent dilemma.
Evidence: In a hypothetical EigenLayer AVS failure, the 10,000 ETH slashing penalty is deducted pro-rata from all restakers in the faulty operator's pool. The individual staker bears the cost for an operator's mistake they did not and could not audit, transforming security into a pooled liability.
future-outlook
THE MISNOMER
Future Outlook: The Security Marketplace
Shared security is a flawed analogy; the restaking economy creates a competitive market for specialized security services.
Shared security is a misnomer. The term implies a communal pool, but EigenLayer and Babylon create a capital allocation market. Node operators bid for work, and protocols pay for bespoke security guarantees.
Security becomes a commodity. The market differentiates on cost, slashing conditions, and validator performance. A Cosmos appchain will not pay the same rate as an Ethereum L2 like Arbitrum.
Protocols demand specialization. An AVS for a fast-finality chain like Solana requires different validator attributes than a slow, data-availability network. Generalized security pools fail this test.
Evidence: EigenLayer's AVS ecosystem already shows this. Omni Network (interoperability) and Lagrange (ZK coprocessor) have distinct technical requirements, forcing node operators to specialize their service offerings.
takeaways
RESTAKING REALITIES
Key Takeaways for Builders and Investors
Shared security is a marketing term; the restaking revolution is about economic bandwidth and risk fragmentation.
01
The Problem: 'Security' is Not Fungible
EigenLayer's pooled security is not a universal good. A validator slashed for an AVS failure loses stake across all services, creating systemic contagion risk. The security of a new chain depends on the weakest AVS in a validator's portfolio.
- Risk: Correlated slashing across unrelated services.
- Reality: Security is service-specific, not a shared pool.
~$20B
TVL at Risk
1->Many
Slash Vector
02
The Solution: Isolate Economic Bandwidth
Projects like Babylon and EigenDA decouple security from validation. They use restaked ETH as cryptoeconomic collateral for specific functions (timestamping, data availability), not for consensus. This isolates risk and creates dedicated security budgets.
- Benefit: Tailored cryptoeconomic security per service.
- Action: Build AVSs that require explicit, isolated stake.
Specialized
Security Budgets
No Contagion
Risk Model
03
The Investor Lens: Yield is Risk Mispricing
AVS rewards are premiums for assuming slashing risk. Current high APYs are a temporary mispricing as the market learns to model complex, correlated slashing events. The end-state is risk-adjusted returns comparable to TradFi.
- Signal: Watch for AVSs with clear, auditable slashing conditions.
- Avoid: Services with vague penalties and opaque operations.
High APY
= High Beta
Risk-Adjusted
True Metric
04
The Builder Mandate: Own Your Security Stack
Relying solely on restaking for security is a critical vulnerability. Follow the Celestia modular playbook: use restaking for specific guarantees (e.g., data availability via EigenDA) while maintaining a dedicated validator set for consensus. Layer security models.
- Strategy: Hybrid models (restaking + PoS).
- Example: A rollup using EigenDA for DA and its own sequencers for execution.
Hybrid
Security Stack
Reduced
Systemic Dep
ENQUIRY
Get In Touch
today.
Our experts will offer a free quote and a 30min call to discuss your project.
NDA Protected
Confidentiality24h Response
Time to ValueDirectly to Engineering Team
No Sales Fluff10+
Protocols Shipped
$20M+
TVL Overall