Put 10 ETH into Lido and you get back about 10 stETH. Look at the wallet a day later and the balance has grown by a few ten-thousandths, without a single transaction on your side. Wrap the same stETH into staking means depositing a proof of stake coin with a protocol that stakes it for you and issues a transferable token (such as Lido stETH or rETH) representing the deposit plus rewards, which you can use or sell while the coins stay staked.">wstETH and the number in the wallet stops moving, but each token is worth a little more stETH every day: as of 29 October 2026, the wstETH contract valued one wstETH at 1.2474 stETH, and Rocket blockchain that runs exactly as written when it is called, holding and moving funds according to its code, without an intermediary able to change the outcome.">smart contract holding reserves of two or more tokens, supplied by users, that a decentralized exchange uses to execute trades automatically at a price set by a formula.">Pool’s contract valued one rETH at 1.1747 ETH. Same staked ether, two ways of keeping the books. That difference sits at the heart of liquid staking, and it decides whether a lending market, a bridge or a tax spreadsheet treats your position correctly.
Behind the tokens sit operators, oracles, queues and a DAO. Lido and Rocket Pool turn deposits into validators in very different ways; stETH traded at a discount in June 2022 because there was then no way to redeem it; Lido’s share of all staked ether still worries Ethereum researchers; and Solana runs the same idea with JitoSOL, mSOL and Sanctum, on a chain where stake unlocks much faster. The previous lesson, on fixed-rate DeFi lending, showed markets where a yield is locked in; a large share of the yield traded in those markets comes from liquid staking tokens, so this is where it starts.
A quick reminder first. Our lesson on interest-bearing tokens explains the two accounting models in general: a rebasing token changes your balance, a value-accruing token keeps your balance and raises the amount each token can be redeemed for. It already walks through the token mechanics of stETH, wstETH and rETH. Here we assume that much and look at the machinery underneath, the numbers as of 29 October 2026 and the risks that come with each design.
Why liquid staking exists: a receipt for ether that is locked
On Ethereum, staking means locking ether in a validator that proposes and checks blocks. A validator traditionally takes 32 ETH, needs a machine online around the clock and faces penalties if it misbehaves or goes offline. Getting in and out is not instant either. On 29 October 2026, validatorqueue.com showed 1.30 million ETH waiting in the entry queue, an estimated wait of 23 days and one hour, and 13,941 ETH in the exit queue, about 21.5 hours. The same page counted 44.2 million ETH staked across 815,671 validators, 36.16% of all ether.
A liquid staking protocol pools deposits of any size, hands them to node operators who run the validators, and gives the depositor a token that represents the claim. That token is the receipt. It can be sent, sold, bridged or posted as collateral while the ether underneath keeps earning consensus rewards and execution-layer fees. Our glossary entry on liquid staking gives the short definition; the rest of this lesson is about what the receipt really promises.
The receipt is what makes the yield usable elsewhere. The stETH and wstETH loop on Aave, explained in our lesson on borrowing against crypto, only exists because the staked ether can be posted as collateral while it is still staked. Restaking, the subject of a later lesson, stacks a further layer on the same receipts.
Rebase or value accrual: stETH, wstETH and rETH
stETH rebases once a day, after an oracle report
stETH does not store a balance for each holder. It stores shares. Your stETH balance is your shares multiplied by the total ether the protocol controls, divided by the total number of shares. When validators earn rewards, the pooled ether grows, the number of shares does not, and every balance rises at once. According to Lido’s tokens integration guide, balances “get recalculated daily” when the protocol’s oracle reports the validators’ balances from the consensus layer.
That oracle is a committee, not a single feed. Lido’s documentation describes nine oracle members, run by node operators the DAO selected, of which five must submit matching reports before an update goes through. Sanity checks cap the reward rate a report may show at 27% a year and reject any report in which staked ether falls by more than 5%. If the committee fails to agree, the rebase waits for the next report, which then covers the missed period. A mass slashing that cut the pooled ether would shrink every balance through the same mechanism: a rebase can go down as well as up, although the guide notes that daily rebases had never been negative when it was written.
Shares have side effects that integrators know well. A rebase changes balances without the token emitting ERC-20 transfer events, so some wallets and accounting tools miss it. Rounding means that sending your whole stETH balance can leave 1 or 2 wei behind on the sending address, a quirk the Lido guide spells out. And any contract that records a stETH balance once, then expects it to stay put, will get its sums wrong.
wstETH and rETH keep the balance and raise the rate
wstETH is the same position written the other way. Lido’s guide defines it as a token that “represents the account’s share of the stETH total supply”. You wrap stETH and receive a fixed number of wstETH; as rewards come in, each wstETH unwraps into more stETH. As of 29 October 2026 the contract’s stEthPerToken function returned 1.24740, which tells you how much the position has grown since wstETH launched.
rETH, Rocket Pool’s token, was built this way from the start: there is no rebasing version. The rETH contract’s exchange rate on 29 October 2026 was 1.17467 ETH per rETH. Market prices sat a hair above both on-chain rates, which is normal for tokens that the market prices slightly ahead of the next oracle update.
| Token | Issuer | Accounting | On-chain rate, 29 Oct 2026 | Price, 29 Oct 2026 (DefiLlama) |
|---|---|---|---|---|
| stETH | Lido | Rebases daily | 1 stETH = claim on 1 ETH | $2,318.74 (1.0003 ETH) |
| wstETH | Lido | Value accrual | 1.24740 stETH | $2,892.94 |
| rETH | Rocket Pool | Value accrual | 1.17467 ETH | $2,724.07 |
| JitoSOL | Jito | Value accrual | 1.30876 SOL | $130.91 |
| mSOL | Marinade | Value accrual | not read | $141.37 |
The ETH price used for the ratios above was $2,317.94 on DefiLlama’s price feed at about 10:00 UTC on 29 October 2026; SOL was near $100.18 at the same time. The JitoSOL rate was read from Jito’s stake pool account on Solana the same day. A stETH price a few hundredths of a percent above or below one ether is market noise, not a premium or a discount.
stETH or wstETH: which one DeFi uses and why
For holding in a wallet, the two are economically the same: one is a wrapper around the other, and you can wrap or unwrap at any time at the contract rate. The difference shows up as soon as the token leaves the wallet. Lido’s guide warns that most cross-chain bridges “have no mechanics to handle rebases” and asks users to bridge only wstETH. That is why the version of Lido’s token you find on Arbitrum, Optimism or Base is wstETH.
Lending markets, vaults and liquidity pools mostly list wstETH for the same reason: a balance that changes on its own breaks the arithmetic of contracts that track deposits. Lido’s own advice to developers is to integrate stETH only when an application deliberately supports its rebasing, and otherwise to prefer wstETH. stETH remains useful where its rebase is wanted, for example in a wallet where you want to see rewards arrive, or in the Curve pool built around it.
For your records, the choice matters too. With stETH, every daily rebase is a small increase in the number of tokens you hold. With wstETH or rETH, nothing arrives; the value of what you already hold grows. Tax authorities have not said clearly whether those two cases are treated the same, a point picked up in the tax section below.
How Lido works behind stETH
Lido takes 10% of staking rewards, according to its documentation, and splits that fee between the node operators and the DAO treasury; the DAO can change the rate. What a holder earns is therefore the protocol’s reward rate minus that 10%. Lido’s API gave a seven-day average of 2.326% on 29 October 2026, while DefiLlama’s yield page showed 2.25% for stETH. Validatorqueue.com estimated the network-wide staking APR at 2.45% the same day, so the gap between running a validator yourself and holding stETH was a fraction of a percentage point, before counting the hardware, the 32 ETH and the time a solo validator needs.
The deposits do not go to one company. Lido’s documentation describes staking modules, each a different way of choosing operators. The Curated Module, now in its second version, serves a vetted set of professional operators: a committee places each one in a group that decides how much stake it receives. The Community Staking Module is permissionless: anyone can join as an operator by posting a bond, and new validators enter through a first-in, first-out queue. Since the latest upgrade, every module requires a bond. It is treated as collateral, and penalties and missing rewards are deducted from it, so the operator has something to lose before stETH holders do.
Since the Shapella upgrade of 12 April 2023, stETH can be turned back into ether through Lido itself. The flow, as Lido’s withdrawal queue documentation describes it, has four steps:
- You approve the contract and call requestWithdrawals with stETH or wstETH, between 100 wei and 1,000 ETH per request.
- You receive an NFT, called unstETH, that represents your place in a first-in, first-out queue and can itself be transferred.
- An oracle report finalises the request once enough ether is available from the buffer or from validator exits and a timelock has passed.
- You claim the ether. The amount is fixed when you request, so rewards earned while waiting are not paid to you.
Lido’s own estimate on the morning of 29 October 2026 was that a 100 stETH request would be finalised in about 50 hours, on 31 October. That figure moves with demand: a crowd of withdrawals that needs fresh validator exits joins the network’s exit queue, which was short that day but has stretched to weeks at busy times. The documentation adds one more condition: if the protocol suffered a large loss on the beacon chain after a request was filed, the request can be finalised at less than one ether per stETH.
Two newer pieces changed what Lido is. Lido V3 went live on Ethereum mainnet, according to a Lido blog post of 30 January 2026, and introduced stVaults, described as “a modular staking primitive designed for stakers that need configurations beyond a single pooled model”. The post lists what a vault can customise: operator selection, policies, fees, reward logic and risk parameters, with optional stETH minted against the vault’s stake. For the rules that keep that minted stETH backed, reserve ratios, risk tiers and forced rebalancing, the post sends readers to the Lido V3 whitepaper.
The other piece is Dual Governance. Lido’s explainer of 4 June 2025 describes a system in which stETH holders can lock their tokens in an escrow to oppose a DAO proposal. Once 1% of stETH supply is locked, a veto-signalling phase extends the delay before execution, from five extra days at 1% up to 45 days at 10%. If 10% is still locked when that delay runs out, a rage-quit state blocks execution until the objecting holders have withdrawn their stake to ether. The aim is that LDO token holders, who run the DAO, cannot push through a change that stETH holders reject without giving them time to leave.
How Rocket Pool and Swell do it differently
Rocket Pool is the usual example of a more decentralised design. Anyone can become an operator by putting up part of the 32 ETH; the protocol fills the rest with ether from rETH holders. Under the Saturn Zero rules in force from 28 October 2024, an operator bonded 8 ETH per minipool and received 24 ETH from the liquid staking side, without having to post RPL, according to saturn.rocketpool.net, an explainer site on Rocket Pool’s domain that calls itself an unofficial guide. The operator’s own ether is the first loss if the validator is penalised.
Saturn One went live on 18 February 2026, according to a Rocket Pool DAO forum update by the core team. The explainer site lists the changes: the bond per validator is halved to 4 ETH, each node deploys one megapool contract that groups several validators and cuts gas costs, deposits enter a standard queue or an express queue, and ether is assigned in two steps (a 1 ETH prestake to hold a place, then the remaining 31 ETH). Protocol rewards are split between rETH holders, operator commission and staked RPL, and the DAO can change that split. The page gives no fixed percentages, and Coindaro has not found a current figure for the commission in Rocket Pool’s documentation.
Early figures show the upgrade working through a crowded network. The same forum update, dated 7 May 2026, reported 60,160 ETH staked through megapools, 18,635 ETH used to mint rETH, and 1,880 megapool validators deposited and waiting in Ethereum’s entry queue. The same post described a hotfix: the activation queue, about 70 days at the time, had grown longer than the window after which a waiting validator would be dissolved. Saturn Two has no fixed date. After a community push to put rETH demand first, its indicative scope includes RPIP-71, rETH withdrawal liquidity using EIP-7002 exits, and RPIP-73, which would exit underperforming validators to protect rETH’s yield.
On 29 October 2026, DefiLlama put Rocket Pool’s total value locked at about $1.13 billion, down about 10% over seven days, against $22.46 billion for Lido. rETH’s yield on DefiLlama’s page was 1.83%, below Lido’s; the share of rewards paid to operators and RPL stakers is the obvious place to look for the difference, though Rocket Pool publishes no single fee figure to compare.
Swell is a smaller example. Its website, read on 29 October 2026, still presented swETH as its liquid staking token for ether, alongside rswETH, a restaking token, and showed a round $200 million for swETH; DefiLlama counted about $28 million in Swell’s liquid staking contracts the same day. A post of 28 April 2026 on the same site announced the sunset of Swellchain, its layer-2 network, with funds to be withdrawn before 15 June 2026, and the home page now says Swell is evolving into Faro, a research-to-trade platform for Hyperliquid. Small LSTs can change course quickly, which is a risk of its own.
Where the yield comes from and how the big LSTs compare
Every ether liquid staking token passes on the same two income streams: consensus rewards paid by the protocol for proposing and attesting, and execution-layer income from priority fees and MEV paid by block builders. What differs is the fee taken, the share of ether kept idle for withdrawals and how well the operators perform. DefiLlama’s liquid staking category listed 300 protocols with $53.1 billion locked on 29 October 2026.
| Ether LST | TVL, 29 Oct 2026 | Yield on DefiLlama, 29 Oct 2026 |
|---|---|---|
| Lido (stETH, wstETH) | $22.46 billion | 2.25% |
| Binance Staked ETH (WBETH) | $8.89 billion | 2.19% |
| ether.fi (weETH) | $4.43 billion | 2.32% |
| Rocket Pool (rETH) | $1.12 billion | 1.83% |
| StakeWise V3 (osETH) | $0.89 billion | 2.20% |
| Coinbase (cbETH) | $0.46 billion | 2.27% |
Lido is about two and a half times the size of the next one, Binance’s token, which is issued by an exchange and depends on that exchange. ether.fi’s weETH is a restaking token, so part of its design belongs to the next layer of risk covered in a later lesson. The yields are DefiLlama’s base figures for that day, not promises; they move every week with network activity.
The stETH discount of June 2022 and what has changed
Our lesson on DeFi composability tells how the stETH discount spread through Aave, Celsius and Three Arrows Capital in 2022. The short version: leveraged positions built on stETH had to be unwound at the worst moment, and the selling pushed the price further down. Here the question is a different one. Why could nobody simply redeem stETH for ether and close the gap?
The sequence, as Fortune summarised a Nansen analysis on 29 June 2022, started with the Terra collapse in May. On 12 May alone, Celsius and Three Arrows Capital were among the wallets that together removed almost $800 million of liquidity from the main Curve stETH/ETH pool. The price relationship broke on 14 May and was not restored that spring; Nansen saw the sharpest deviation between 9 and 11 June. Celsius froze withdrawals on 12 June, and Three Arrows only began unwinding its stETH positions on 13 and 14 June. Reports at the time put the discount at about 5% to 6%, a rate of roughly 0.94 ETH per stETH at the low.
The answer to the redemption question is simple: there was no way out. Before Shapella, ether staked on the beacon chain could not be withdrawn at all, by Lido or anyone else. stETH was a claim on ether that would come back on a date nobody knew. A holder who needed ether that week had one exit, a sale on a decentralised exchange, at whatever price buyers offered. The discount measured how much the market charged for waiting an unknown time.
Withdrawals were enabled with the Shanghai and Capella upgrade on 12 April 2023. Since then the price of stETH has an anchor. If it trades well below one ether, anyone can buy it cheaply, queue it for withdrawal and collect a full ether a few days later. On 29 October 2026 stETH traded within a few hundredths of a percent of one ether on DefiLlama’s prices, and Lido estimated about two days for a 100 stETH withdrawal.
An anchor is not a guarantee. A discount can still open for anyone who needs ether faster than the queue, because waiting costs something. In a rush, the network’s exit queue lengthens, and the queue it takes to get out of Lido stretches with it. A slashing event or a bug in the contracts would cut the ether behind every token, since losses are shared across all holders. And a lending market that liquidates on the market price, not the redemption rate, can sell positions during a short dip even when the ether is all there.
Lido’s share of staked ETH and the centralisation debate
How much of all staked ether does Lido control? There is no single official figure. Taking Lido’s own count of 9.69 million pooled ETH and validatorqueue.com’s total of 44.2 million ETH staked, both read on 29 October 2026, gives about 22%. That is Coindaro’s own division of two figures from different sources, not a dashboard reading, and other trackers that count pending deposits differently will show a slightly different number. In June 2022, The Block reported that Lido accounted for “almost a third” of ether staked on the beacon chain.
Those percentages matter because Ethereum’s proof of stake has thresholds. Ethereum.org’s page on attacks and defences describes what a single party controlling a given share of stake could do:
| Share of staked ETH | What a single controller could do, per ethereum.org |
|---|---|
| 33% | Prevent the chain from finalising by going offline together; the inactivity leak then drains that stake until finality resumes |
| 34% | Attempt double finality, a lasting chain split, at the cost of its entire stake through slashing; needs control over message timing |
| About 50% | Split the chain into two forks and block finality until the inactivity leak lets both forks finalise, leaving the community to choose one |
| 66% | Finalise the chain it prefers without help from honest validators, control what goes into finalised blocks and censor transactions |
Lido is not one validator. Its stake runs on many independent operators across its modules, and none of them comes close to those shares alone. The worry is the layer above. One DAO and its committees decide who joins the curated set, one set of contracts receives every withdrawal, and one token, LDO, governs both. If that governance were captured or coerced, the operators would follow the rules it sets. Ethereum.org’s page on pooled staking warns that large amounts of staked ETH under a few organisations “create conditions for censorship”, and tells readers to prefer pools with permissionless, distributed operator sets.
Lido’s community was asked the question directly. A vote that opened on 24 June 2022 asked whether Lido should limit its share of the liquid staking market. With about a day of voting left, The Block counted more than 99.8% of votes against any self-limit. Instead of a cap, Lido later added the permissionless Community Staking Module, Dual Governance and stVaults. Rocket Pool answers the same problem differently, with an operator set open to anyone and a bond it has cut from 16 ETH at launch in 2021 to 4 ETH.
The share has fallen since 2022 as exchanges, ether.fi and others grew, and Lido now sits below the 33% line by a clear margin. The debate has not gone away: it now turns on who controls the contracts, and on how fast a share can rebound when yields change.
Liquid staking on Solana: JitoSOL, mSOL and Sanctum
On Solana, staking works through stake accounts delegated to validators, and liquid staking runs mostly on the shared stake pool program: a pool holds a list of validators and a reserve, and depositors mint pool tokens at the current exchange rate. All the main Solana LSTs accrue value, like wstETH and rETH; none of them rebases. On 29 October 2026, Solana’s public RPC reported about 433 million SOL staked with 665 active vote accounts.
JitoSOL is the largest single-pool token. Reading Jito’s stake pool account on 29 October 2026 gave 10.42 million SOL held against 7.96 million JitoSOL, an exchange rate of 1.30876 SOL per JitoSOL, or about 2.4% of all staked SOL by Coindaro’s own calculation. Jito’s documentation says JitoSOL “automatically accrues value from staking and MEV rewards”. The pool stakes only with validators running Jito’s software, which collects tips that traders pay to have their transactions placed in a block, and the MEV share that the pool receives reaches holders through the exchange rate. Jito’s documentation page did not state the pool’s fee, so we do not quote one.
Marinade runs three products, according to its documentation. Marinade Liquid issues mSOL, which grows in value with rewards and charges, in the docs’ words, “no performance fee on staking rewards”; a delayed unstake carries a 0.2% protocol fee. Marinade Native leaves stake accounts in the user’s own wallet and delegates them through an auction among validators. Marinade Select, a curated validator set, was not accepting new SOL when the documentation was read.
Sanctum takes another route: instead of one big pool, it is infrastructure on which validators and apps launch their own LSTs. Its site invites validators and apps to join Jupiter, Bybit and Drift in launching white-label tokens, and a shared liquidity pool, Sanctum Infinity, lets holders swap between these tokens or back to SOL; DefiLlama counted about $193 million in it on 29 October 2026. DefiLlama groups them as Sanctum Validator LSTs.
| Solana LST | TVL, 29 Oct 2026 (DefiLlama) | Yield on DefiLlama, 29 Oct 2026 |
|---|---|---|
| Sanctum Validator LSTs (many tokens) | $1.77 billion | varies by token |
| Binance Staked SOL (BNSOL) | $1.08 billion | 4.54% |
| Jito (JitoSOL) | $1.04 billion | 4.90% |
| Jupiter Staked SOL (JupSOL) | $0.52 billion | 5.09% |
| Drift Staked SOL | $0.28 billion | not listed |
| Marinade (mSOL) | $0.23 billion | 5.41% |
Exiting is also different. There is no multi-week validator queue on Solana; a deactivated stake account becomes withdrawable when the epoch ends. An epoch is 432,000 slots; since Solana shortened its slot time in the autumn of 2026, the last eleven epochs before 29 October each lasted about 1.07 days, by Coindaro’s reading of block times on the public RPC, against roughly two days earlier. In practice most holders who want SOL at once swap through a liquidity pool, at a small fee or a small discount. The same rule applies as on Ethereum: the market price can drift from the exchange rate when many people want out at the same time. Lido, once present on Solana, now shows only about $5 million there on DefiLlama, a negligible amount.
Risks to check before holding a liquid staking token
None of these risks is hidden, but they are easy to forget once a token sits in a wallet and its number goes up. Six checks before holding an LST or using it in DeFi:
- The staking contracts, the withdrawal queue and any wrapper can contain bugs, and a loss would be shared by all holders.
- The token trades on markets and can sit below its redemption value, especially when the exit queue is long.
- A misbehaving or badly run operator can be slashed, and the operator’s bond only covers a first slice of that loss.
- A DAO can change fees, operators or contracts, so it is worth knowing who votes and what holders can do if they disagree.
- Bridging stETH instead of wstETH, or sending a rebasing token to a contract that does not support rebases, can cost you the rewards or the tokens.
- Looping an LST against borrowed ether multiplies each of the risks above and adds liquidation, as our lesson on borrowing against crypto explains.
The flow can be rehearsed without real money. Lido’s integration guide says the most recent testnet version of the protocol runs on Hoodi, an Ethereum test network, and Lido’s list of Hoodi contracts includes stETH, wstETH and the withdrawal queue, so wrapping, unwrapping and queueing a withdrawal can be tried there with free test ether first. On mainnet, each of those steps is a separate transaction that costs gas, which weighs more on a small position than on a large one.
Taxes on liquid staking in the US and the UK
In the United States, the IRS published Revenue Ruling 2023-14 in 2023. It holds that the fair market value of staking rewards is included in gross income in the year the taxpayer “gains dominion and control” over them, valued at that moment, including when staking goes through an exchange. The ruling does not address liquid staking tokens. Whether swapping ETH for stETH is a disposal, and whether a stETH rebase is income on the day it arrives while wstETH’s rising rate is only taxed when you sell, are not settled by IRS guidance. Our guide to crypto taxes in the United States covers the general rules.
In the United Kingdom, HMRC deals with the question in its Cryptoassets Manual, in the decentralised finance pages on lending and staking (CRYPTO61000 onwards). HMRC notes that the terms “lending” and “staking” have no statutory or legal meaning in this context, and asks whether beneficial ownership of the tokens passes when they are deposited; if it does, the deposit is a disposal for capital gains tax. The manual does not name liquid staking tokens, so the analysis depends on the facts of each protocol. Our guide to crypto taxes in the United Kingdom has the wider picture. In both countries, keeping a record of every deposit, wrap, unwrap and withdrawal, with dates and dollar or pound values, makes any later answer easier to apply.
The next lesson of the DeFi path turns to restaking: what happens when the same staked ether is pledged a second time to secure other services, and why the risks stack.
