Liquid Staking vs Native Staking: Who Runs the Validator

Two different objects, not two settings

Native staking, which ethereum.org calls home staking, is a protocol role. You run an Ethereum node, deposit at least 32 ETH to activate a validator, hold the validator keys, and are paid by the protocol. The staking page describes it as the baseline: “Nothing stands between you and the protocol: you hold your own keys, and the protocol pays full participation rewards directly to you.”

Liquid staking is a claim on that role, held by someone else. The pools page describes staking pools as a way for people with less than 32 ETH to combine deposits, and says that some pools “operate using smart contracts, where funds are deposited to a contract that manages and tracks your stake, and issues you a receipt token (liquid staking token) that represents this value.” The protocol stakes the ETH with the pool’s node operators; you hold the token.

The separating sentence is on the same page: “The Ethereum protocol pays rewards to validators; it doesn’t know your token exists.” When you hold a liquid staking token, you are not a staker from the protocol’s point of view.

The mechanics on each side

Native staking. A validator needs at least 32 ETH. With regular (0x01) withdrawal credentials its effective balance is capped at 32 ETH and any excess is swept to your withdrawal address every few days; with compounding (0x02) credentials the effective balance can grow to 2,048 ETH, rewards compound automatically, and withdrawing anything below the cap means triggering a partial withdrawal from the execution layer, which costs gas. The deposit is recognized by the network in around 13 minutes, but new validators pass through an activation queue whose length varies with demand. Exiting means submitting a withdrawal and waiting for a queue that depends on network demand.

Two operational details matter. The withdrawal address can be set only once and cannot be changed afterwards. And slashing is not downtime: provable misbehavior destroys part of the stake and forcibly removes the validator, while being offline costs small inactivity penalties.

Liquid staking. You deposit ETH into the protocol’s contracts, the protocol stakes it with its node operators, and you receive the receipt token. Rewards show up in one of two designs, as the pools page sets out: rebasing tokens, where your balance grows so one token stays roughly equal to one ETH, and exchange-rate tokens, where the balance stays fixed and each token becomes redeemable for more ETH over time. Both deliver rewards net of the protocol’s fee.

The exit has two paths. Redeeming through the protocol depends on liquidity being available, either an unstaked buffer or validators exiting through the consensus layer queue, which can take time. Selling on a secondary market is immediate but the price can deviate from the backing value.

A worked example: what the receipt token represents

Every number below is illustrative; none is a rate you can get.

Suppose an exchange-rate token redeems for 1.05 ETH and you deposit 10 ETH.

• tokens received = 10 ÷ 1.05 = 9.5238095238
• if the redemption rate later reaches 1.10 ETH per token, the position is worth 9.5238095238 × 1.10 = 10.4761904762 ETH
• gain = 10.4761904762 − 10 = 0.4761904762 ETH, or 4.7619% of the deposit

Now suppose you need out while the secondary market pays 1.00 ETH per token:

• proceeds = 9.5238095238 × 1.00 = 9.5238095238 ETH
• shortfall against redemption value = 10.4761904762 − 9.5238095238 = 0.9523809524 ETH
• as a percentage: 9.5238095238 ÷ 10.4761904762 − 1 = −9.09%

The token did its job as a transferable claim and still delivered 9.09% less than the redemption value, because the market price, not the protocol, set the exit. A rebasing version of the same position would show the identical economics in a different place: instead of the per-token rate rising from 1.05 to 1.10, the balance would rise from 10 to 10.4761904762.

Where the naive reading goes wrong

“Liquid staking is native staking with a token attached.” The protocol pays validators, not token holders. Holding the token means holding a claim on a protocol that stakes on your behalf, and the pools page lists what that adds: smart-contract risk, market and liquidity risk, governance and upgrade risk, operator-set centralization, and slashing pass-through.

“The token is always redeemable at the underlying rate.” Redemption depends on the protocol having liquidity, either a buffer or validators that can exit through the consensus queue. The page says a secondary-market sale at a discount may be the only fast exit when redemptions are congested.

“The yields are directly comparable.” A liquid staking return is net of the protocol’s fee, and the two token designs express the same return differently: a rebasing balance and an exchange-rate token can hold identical value and look nothing alike.

“Slashing is the operator’s problem.” The pools page states that if the pool’s validators are slashed, the loss is typically socialized across all token holders under the protocol’s rules. You are exposed to operators you did not choose.

“A higher yield is just better staking.” The page warns that products advertising enhanced yield may combine staking with restaking, which commits the ETH to additional protocols under additional slashing conditions, and suggests asking where any yield above the core network rate comes from.

What the primary sources do and do not say

The staking landing page gives the 32 ETH minimum, the 2,048 ETH maximum, and the definition of slashing as a penalty for provable misbehavior in which part of the stake is destroyed and the validator is removed. It also displays a “Current APR” figure under a page stamp of February 12, 2025 when retrieved; because that figure is not dated to today, this article quotes no yield.

The home-staking page supplies the deposit recognition time, the activation queue, the two credential types, the exit path, and the distinction between inactivity penalties and slashing.

The pools page supplies the receipt-token mechanics, the rebasing and exchange-rate designs, the redemption and secondary-market exits, the risk list, and the statement that liquid staking protocols account for around a third of all staked ETH, on a page last updated August 17, 2026. It lists protocols as examples, which the page says is not an endorsement.

The withdrawals page supplies the rule that the withdrawal address is set once and cannot be changed, and that a full exit waits for a queue that depends on network demand.

None of the four pages states a slashing percentage, a typical queue duration, or a price for any token, so none of those figures appears here.

A checklist before you choose

1. Name the key holders. For native staking, that is you: signing keys on one machine, a minority client, and a withdrawal address checked before submission because it cannot be changed later.
2. For a liquid token, verify the claim. Can you see onchain where the ETH goes, is the operator set published, and is the token in your own wallet and redeemable? The pools page poses these as the test of transparency.
3. Identify the reward design. Rebasing or exchange-rate changes how the position appears and how it behaves in other applications.
4. Price the exit. Know whether you would redeem through the protocol or sell, and what each path depends on.
5. Treat extra yield as a question. If a product pays materially more than core staking, find out which additional slashing conditions are funding it.

How these sources were checked

Ethereum.org, “Ethereum staking,” retrieved 2026-10-23. Ethereum.org, “Home stake your ETH,” retrieved 2026-10-23. Ethereum.org, “Liquid and pooled staking,” retrieved 2026-10-23. Ethereum.org, “Staking withdrawals,” retrieved 2026-10-23.