Liquid Staking and Consensus Risk: Centralization Concerns

The liquid staking consensus centralization risk emerges when a few liquid staking protocols (like Lido on Ethereum) accumulate >50% of a network’s stake, giving their operators outsized influence over which blocks are produced and which validators earn rewards—a structural threat to decentralization that may be as serious as a 51% attack.

The structural problem: minimum stake and pooling

Ethereum requires 32 ETH (~$100K) to run a solo validator. Most retail investors cannot afford this. Instead, they deposit into a liquid staking protocol (LST), which pools capital, runs validators, and returns a liquid token (e.g., stETH for Lido) that can be traded or used as collateral.

This is convenient: an investor with 0.1 ETH can stake and receive 0.1 stETH, earning staking yield immediately. But it is also dangerous: the investor’s stake is now controlled by the LST’s operators. Those operators can choose which validators to back, how rewards are distributed, and how network fees are set. If Lido’s operators decided tomorrow to use their influence to censor transactions on Ethereum, there is no protocol-level recourse—they have the votes.

This is not speculative: Lido controls ~32% of Ethereum’s stake as of 2025, and the top five LST protocols control >60%. At this level, an attacker who compromises Lido’s servers could halt the network without slashing, because Lido’s 32% of validators would go offline. Worse, Lido’s team could voluntarily coordinate a censorship campaign (censoring certain transactions or validators), and Ethereum’s protocol has no way to force them to include those transactions.

How concentration enables attacks

A liquid staking operator who controls C% of a network’s stake can:

1. Censor transactions: refuse to include them in proposals, effectively freezing funds or excluding users.
2. Block particular validators: by controlling which validators receive delegation, they can force certain node operators out of the set.
3. Coordinate slashing: propose equivocations (conflicting blocks) to slash competing validators, then absorb the slashing penalty themselves (e.g., if they hold 60% of stake and slash themselves, the protocol burns 60% of the penalty, but the victim is gutted).
4. Halt the network: if online for longer than the inactive leak threshold, offline validators lose stake over time; a concentrated operator could simply shut down servers, forcing the network to depend on other validators (which reduces throughput and confirms liveness relies on a handful).

None of these require >50% stake in the traditional “51% attack” sense. A 33% player can slow finality, because the protocol waits for 2/3 supermajority votes; if they withhold, finality stalls. A 50% player can outright fork. But even smaller players can cause asymmetric harm: censorship requires only control of block proposals, not consensus.

Lido’s dominance on Ethereum

Lido concentrated Ethereum’s staking because it solved the entry problem with a superior product:

• No minimum: deposit any amount.
• Liquid token: stETH can be traded, lent, or used in DeFi, earning additional yield.
• Professional operation: Lido’s node operators (Coinbase, Figment, Blox, etc.) are well-resourced and reliable.

By 2024–2025, Lido reached 32% of Ethereum’s stake. The runner-up, Coinbase, has ~13%. After that, Rocket Pool, Kraken staking, and others are much smaller. In effect, Ethereum’s consensus is betting on Lido not being compromised, censoring, or making mistakes.

The protocol has no formal stake cap (unlike Polkadot, which uses nominator caps per validator to limit pool concentration). Some developers proposed a mechanism to penalise Lido if it exceeds 22% (a Lido operator called it the “Lido Limit”); it was never activated, citing fairness and complexity.

The liquid staking token’s role

A liquid staking token (stETH, rETH, etc.) is itself a counterparty risk. If Lido’s smart contract is hacked, users’ deposits can be stolen. If Lido’s operators go rogue, they control the withdrawal queue and can freeze exits. If Lido’s token de-pegs from ETH (loses its 1:1 peg), users face a liquidity crisis: they cannot convert stETH back to ETH at par, and selling stETH in the open market incurs slippage.

A de-peg cascade can trigger a run: as stETH trades below 1.0 ETH, nervous holders rush to exit the staking contract or dump stETH on exchanges. If the backlog of exits exceeds Lido’s reserves (ETH available to withdraw), users face indefinite queuing. This can force staking withdrawals offline (Lido shuts down validators to conserve ETH), which halts Ethereum’s consensus.

This happened in miniature during crypto crashes (e.g., FTX collapse, 2022): stETH hit 0.96 ETH on secondary markets, causing panic. Though it recovered, the vulnerability was clear.

Systemic risk: interconnectedness

Liquid staking tokens are now widely used as collateral in DeFi—borrowed against, used in yield strategies, and held in money market protocols. A Lido de-peg would trigger cascading liquidations: lenders would seize stETH collateral, flooding the market and deepening the peg loss. Simultaneously, Ethereum’s consensus would fracture (if Lido offline) or face concentrated control. This is a systemic failure: crypto-wide financial contagion + blockchain consensus breakdown.

Regulatory risk adds a layer: if governments order Lido (a company incorporated in multiple jurisdictions) to censor certain transactions, it must comply or risk seizure. Users have no choice—their stETH is trapped in Lido’s system.

Other networks’ liquid staking risks

Polkadot uses a different model: nominators delegate to validators, not to pools directly, and the protocol itself limits any single validator’s share of active set via dynamic list capping. This limits concentration but has its own problems (reduced capital efficiency, smaller validators can’t compete).

Cosmos Hub has seen Lido introduce stLUNA and stCosmos products, creating the same concentration risk. Larger Cosmos chains (Osmosis, Evmos) have their own LSTs. The fragmentation reduces Cosmos Hub’s security as capital flows to higher-yield chains.

Solana is even more concentrated: ~50% of Solana’s stake flows through a handful of exchanges (Kraken, Coinbase) and validator companies. Solana’s lower minimum stake has not solved concentration; it is just that the barriers are capital (to run high-performance validators) rather than token supply.

Proposed defenses

Staking caps: Limit any single operator to X% of stake. Ethereum developers discussed capping Lido at 22%; Polkadot enforces per-nominator caps. Drawbacks: hard to enforce (operators run hidden validators), and caps discourage the best operators from growing.

Dispersal incentives: Reward validators who have lower concentration, or penalise those with high concentration. This is economically soft and can be gamed (operators spin up dummy validators).

Validator exit delays: If Lido wants to withdraw stake after misbehaving, force a long unbonding period (Ethereum has 27 hours; Polkadot has 28 days). This discourages exits but doesn’t prevent concentration.

Solo staking improvements: Lower the minimum stake or improve solo staking UX to reduce LST appeal. Ethereum considered this; the trade-off is higher state overhead (more validators = more work for nodes). Some proposed “enshrinement”—baking liquid staking into the protocol itself so no single operator controls it.

Slashing transparency: If concentrated operators know slashing is swift and public, they behave better. But slashing deters honest operators too, potentially accelerating concentration (only big, well-funded firms can absorb slashing risk).

Economic incentives at odds

Here is the paradox: the minimum stake was designed to promote decentralization by requiring attackers to lock large capital. But it did the opposite—it created barriers that forced most users into pools, which consolidated power. A lower minimum would give more users solo staking options, but it bloats the validator set and increases node overhead, potentially centralising things differently (only well-funded nodes can afford to run).

No easy answer exists. Some researchers argue that LSTs are inevitable and protocols should design for it (e.g., “liquid democracy” in consensus design, where stakers vote on governance issues, not just attest to blocks). Others argue for stronger protocol mechanisms to break concentration.

Real-world consequences

In practice, Lido has not censored transactions or coordinated slashing. But in 2023, Lido’s team paused new validator deployments after criticism, showing they are responsive to community pressure. This is good governance in the moment but also highlights the vulnerability: consensus power is held at the sufferance of a business, not by protocol rule. If Lido was acquired by an authoritarian regime, sold, or suffered key-person risk, Ethereum could find its consensus broken overnight.

See also