Liquid staking and what a staking receipt is really worth

Liquid staking solves a specific and unglamorous problem. Running an Ethereum validator requires 32 ETH, technical uptime and a willingness to wait in a queue to get out again. A liquid staking protocol pools deposits from anyone with as little as about 0.01 ETH, runs the validators on their behalf, and issues a token that represents the deposit plus accrued rewards. That token is transferable. You can hold it, sell it, lend it or post it as collateral while the ETH underneath stays staked and keeps earning.

Nothing about the underlying reward changes. The network paid an APR of2.46% across 911,289 active validators and43.1 million ETH — about 35.31% of supply — when we checked. A liquid staking token pays that, minus the provider's cut. What you are buying is liquidity and a lower minimum, not a better rate. What you are taking on is a smart contract, a fee, and the possibility that the receipt trades below the thing it represents.

Rebasing tokens and rate-based tokens

Every liquid staking token has to answer the same question: where do the accrued rewards actually show up? There are two designs and they are not interchangeable in practice.

A rebasing token keeps the price anchored and grows the balance. Lido's stETH works this way: one stETH stays approximately one ETH, and the number in your wallet increases daily as rewards land. It is intuitive to read and awkward to integrate, because a great many DeFi contracts assume balances only change when someone moves them. It is also awkward for tax and accounting systems that expect discrete receipts.

A rate-based token does the reverse: the balance never changes and the exchange rate to ETH climbs. Lido's wstETH is simply stETH wrapped into this form, and you can move between them freely. Coinbase's cbETH and Binance's WBETH are natively rate-based. One WBETH represents one staked ETH plus everything it has earned since27 April 2023 at 08:00 UTC, with the ratio updated daily. Binance's older BETH behaves more like a rebasing token, distributing rewards to the spot wallet daily from T+2. Wrapping between WBETH and BETH carries no fee on Binance, and on-chain redemption runs through Binance's own contract.

Who actually runs this market

The liquid staking sector held roughly $50.26bn across 276 protocols, of which the Ethereum LST subtotal was 17.19 million ETH, about $41.25bn — meaning roughly 40% of all staked ETH is liquid-staked. Concentration within that is severe, and it is the single most cited criticism of the model.

Two things in that table deserve more than a glance. The first is the net APY column: every major provider lands between 2.05% and 2.73%, a spread of 30 to 50 basis points across the whole group. That is not a performance ranking. It is a fee schedule wearing a performance ranking's clothes, and it means the sensible way to choose between them is on custody model, decentralisation and integration depth rather than on the third decimal place of a yield.

The second is the cbETH row. Two reputable trackers give Coinbase's commission as 25% and 35%, Coinbase's own cbETH page states no commission at all and notes that cbETH itself cannot be staked on Coinbase, and the reported TVL differs by a factor of nearly three between DefiLlama and Staking Rewards because they are measuring different bases. We could not resolve any of it, so we print the conflict rather than picking a side.

Pegs, and what a discount actually tells you

The recurring anxiety about receipt tokens is that one of them stops being worth one ETH. It has happened twice on a scale worth studying, and the two episodes have almost nothing in common.

The generalisable lesson is that a post-Shapella LST discount is a liquidity event rather than a credit event. It is produced by leverage loops unwinding into a congested exit queue, and its size tracks how long the queue is. That makes the exit queue, not the protocol's reserves, the number to watch. It also means the risk is real but bounded and mean-reverting, which is a very different proposition from the 2022 episode where there was simply no exit.

Rocket Pool and the case for a smaller provider

Rocket Pool holds 3.04% of the LST market against Lido's 56.53%, and it exists largely because of that gap. Its node operators post their own capital alongside pooled ETH, which means the validator set is permissionless rather than curated. Saturn I went live on18 February 2026 and changed three things at once: the operator bond fell from 8 ETH to 4 ETH per validator, megapools now group minipools under a single contract for gas savings, and RPL staking became optional while earningETH from protocol commission rather than RPL inflation. The split of commission between rETH holders, node operators and RPL is governance-adjustable.

We should be straightforward about a limit here: Rocket Pool's site renders entirely in JavaScript and its documentation homepage exposes no figures, so we could not verify current headline commission percentages from primary documentation. The 5–14% range in the table above comes from third-party trackers. That is a transparency finding, not an accusation — but on a page about fees, a provider whose fee is not readable without running a client-side application is worth noting.

Restaking: what it was, and what happened to it

Restaking lets already-staked ETH, or a liquid staking token representing it, be pledged a second time as cryptoeconomic security for third-party services — data availability layers, oracles, bridges, co-processors, collectively called AVSs. In exchange for extra fees, you accept additional slashing conditions written by each of those services. A liquid restaking token then wraps that position into yet another transferable receipt, so the stack becomes LST contract, LRT contract, restaking contract, AVS.

The risks that come with that are not subtle: a second, non-Ethereum slashing surface defined by parties with no obligation to you; operator concentration, since a small set of operators controls most restaked capital; stacked smart-contract exposure across four layers; depeg risk on the LRT itself; AVS governance that can change the conditions under you; and — historically the decisive one — yields driven by points and airdrops rather than by real AVS fee revenue.

The path there is worth knowing in outline. On 24 April 2024 Renzo's ezETH briefly printed around $700 on Uniswap after a disappointing REZ tokenomics disclosure collided with Binance launchpool farmers receiving tokens two days early. That single hour produced more than $65m of liquidations on Morpho and Gearbox, leaving $34k and $83k of protocol bad debt respectively. It demonstrated that thin-AMM restaking tokens are far more fragile than the LSTs underneath them, because there is no redemption arbitrage holding the price up in the minutes that matter.

EigenLayer's own slashing went live on mainnet on 17 April 2025, opt-in per operator and staker — which is to say the risk everyone had been pricing since 2023 only became real two years later. By 6 August 2026 ether.fi had removed restaking from weETH entirely, moving it into a separate opt-in token built on Symbiotic rather than EigenLayer, with under 1% of its assets still restaked against roughly 50% at the start of the year. EigenLayer has repositioned as EigenCloud, selling EigenDA, EigenAI and EigenCompute; EIGEN had about 740 million tokens circulating of 1.82 billion total and a market capitalisation near $145m as of 16 May 2026.

Where liquid staking fits

Strip away the token mechanics and liquid staking is a fairly simple trade. You give up roughly 10% of your staking rewards, about 25 basis points a year, and accept smart-contract exposure. In return you get a 0.01 ETH minimum instead of 32, an exit through the market rather than the queue, and a token that other protocols will accept as collateral. For a holder who wants staking exposure without running infrastructure, that is a reasonable exchange and the fee is a great deal lower than most custodial alternatives — Kraken takes 25% to 30% depending on product and balance, and Coinbase's standard commission is around 35%.

What it is not is a way to earn more. It is also not a substitute for understanding the base layer: the network APR, the entry queue, the issuance formula and the slashing rules all still apply to the ETH underneath, and they are covered inEthereum staking and crypto staking. If you are comparing the receipt-token approach against a custodial savings product, the trade-offs are laid out in staking vs savings, and the cross-asset rates in staking rewards compared. For what the same idea looks like on Bitcoin — where there is no native staking to wrap in the first place — see Bitcoin yield and BTCfi. And if you want the composability rather than the staking, DeFi yield farmingexplains what people actually do with these tokens once they hold them.