Liquid Staking Explained: How It Works, Yield, and Risk

Staking secures Ethereum and pays a reward for doing it, but the classic version locks your coins away, asks for a 32 ETH deposit, and expects you to run a server that never sleeps. Liquid staking removed all three walls. You deposit ETH, you receive a token that keeps earning the staking reward, and that token stays free to trade, lend, or post as collateral while the ETH underneath it keeps validating the network. One idea, and it now sits at the center of Ethereum’s economy: about a third of all ETH is staked, and liquid staking tokens make up the single largest slice of it.

This guide covers what liquid staking is, how the tokens actually behave, where the yield comes from, who the major players are in 2026, and the risks that never appear on a marketing page. It also walks through the two developments that reshaped the sector this year: a validator consolidation that is quietly rebuilding Ethereum’s node set, and a contested proposal to burn staking rewards outright once too much ETH is locked up.

What Liquid Staking Is, and Why It Exists

Ethereum runs on proof of stake. Instead of miners burning electricity, validators post ETH as a bond and earn rewards for proposing and attesting to blocks; misbehave or go offline for long stretches, and part of that bond can be slashed or bled away. To run a validator directly you need exactly 32 ETH, hardware that stays online around the clock, and the discipline to keep the software patched. That is a high bar, and until the Shapella upgrade in April 2023 the staked ETH could not even be withdrawn.

Liquid staking dissolves the barrier. A protocol or a company pools deposits from many users, runs the validators on their behalf, and hands each depositor a liquid staking token (an LST) that represents their share of the staked ETH plus the rewards accruing to it. You can hold any amount, you do not touch a server, and, crucially, the receipt token can move: you can sell it, swap it, or lend it while the ETH behind it keeps working. Lido, the protocol that turned the idea into a mass product, launched in December 2020, back when staked ETH was frozen with no exit at all, so the receipt token was the only way to pull liquidity out of a deposit. The name says it plainly. Staking that stays liquid.

The demand for that liquidity is easy to see in the network’s own plumbing. As of late August 2026 roughly 42.5 million ETH is staked, about 34.8 percent of the supply, spread across 902,651 validators, with a base reward near 2.65 percent, according to validatorqueue.com. The catch is the queue: new validators wait in an entry line that recently held about 2.13 million ETH and stretched past 37 days. A liquid staking token skips that line entirely, because buying or selling one on the open market never creates or destroys a validator. That is the core convenience, and it is why LSTs became the default way most people stake.

How a Liquid Staking Token Actually Works

The lifecycle has two doors. Through the front door you deposit ETH into the protocol’s smart contract, the contract routes it to professional node operators who spin up validators, and you receive freshly minted LSTs in return. To exit through that same door you burn the token and join a withdrawal queue that unwinds a validator and returns your ETH, a path that only exists because Shapella switched withdrawals on. Through the side door you simply buy or sell the token on a decentralized exchange such as Curve or Uniswap, or on a centralized venue, which settles instantly and never touches a validator at all.

Who runs the validators is where the designs diverge. Lido, the largest protocol, uses a curated set of around three dozen vetted operators plus a permissionless Community Staking Module for solo participants. Rocket Pool takes a permissionless route from the start, letting anyone become a node operator by posting a bond alongside pooled ETH from rETH holders. The operators earn a commission, and the protocol takes a fee on top: Lido, for example, charges 10 percent of staking rewards, split evenly between operators and its DAO treasury, per lido.fi. What you hold is a claim, enforced by code, on the ETH those operators are validating and the rewards it produces.

Rebasing or Reward-Bearing: stETH, wstETH, and rETH

Every LST has to answer one question: how does it show you the rewards you are earning? There are two answers, and knowing which one you hold matters for taxes, for DeFi, and for reading the price.

A rebasing token, of which Lido’s stETH is the flagship, keeps a price of roughly one ETH and pays you by increasing your balance a little each day. Open your wallet tomorrow and you own slightly more stETH than today. It reads intuitively, but a moving balance breaks some smart contracts and raises awkward questions about when a reward becomes taxable. A reward-bearing token, sometimes called an exchange-rate token, fixes the number of tokens you hold and lets each one grow more valuable against ETH instead. Rocket Pool’s rETH is worth about 1.1719 ETH in late August 2026, and wstETH, the wrapped form of stETH, trades around 1.24 ETH, per CoinGecko. That higher price is not a premium and it is not a depeg; it is simply two and a half years of accumulated staking rewards baked into one token.

The practical takeaway: if you plan to use your LST inside DeFi, you almost always want the reward-bearing version, which is why wstETH, not stETH, is the form most lending markets accept as collateral.

Where the Staking Yield Actually Comes From

Staking yield is not conjured from a token printer; it is paid for real work securing the chain, which is what makes it a form of real yield rather than emissions. It arrives in three streams. The first is consensus issuance, the new ETH the protocol mints to reward validators for attesting and proposing, currently worth about 2.65 percent a year per ethereum.org. The second is priority fees, the tips users attach to transactions to jump the queue. The third is MEV, the value captured from ordering transactions within a block, usually routed to validators through MEV-Boost. Stack the three and a validator earns somewhere near 3 to 4 percent all in, before the protocol’s cut. The mechanics of that reward stack, and how operators actually get paid, are broken down in our guide to validator economics.

That number has been sliding. In mid-2023 the base reward topped 5 percent; it has drifted toward 2.65 percent since, and the reason is arithmetic. Ethereum’s issuance curve pays out a roughly fixed pool of new ETH that gets divided among all validators, so the more ETH that piles into staking, the thinner each slice becomes. Liquid staking, by making it trivial to stake, is part of what drove the staking ratio up and the per-validator reward down. Hold that thought, because it is exactly the dynamic behind the most contentious Ethereum proposal of the year.

The 2026 Market Map: Who Holds What

Liquid staking is a concentrated market, and one name towers over it. Lido’s stETH carries a market value around 23.4 billion dollars across 9.63 million tokens, per CoinGecko. Within the pure liquid staking segment Lido controls roughly 62 percent, about 8.89 million ETH, while accounting for something closer to 23 percent of all staked ETH once you count exchanges and solo validators, according to DefiLlama and Datawallet. The broader liquid staking category tracked by DefiLlama spans more than 270 protocols and over 50 billion dollars in value, but the long tail is thin.

The surprise for many readers is the runner-up. The second-largest liquid staking token is not a DeFi darling but Binance’s wBETH, worth roughly 9 billion dollars, a top-20 crypto asset in its own right per CoinGecko. Rocket Pool’s rETH sits near 910 million dollars, Coinbase’s cbETH has shrunk below 500 million and keeps contracting, and Frax’s two-token frxETH system holds a few hundred million more. The table below sketches the field.

Read the map and a pattern jumps out: two of the five biggest liquid staking tokens are run by centralized exchanges. That is not a footnote. It defines the most important choice a staker makes.

Custodial or Decentralized: Two Different Trust Models

Two liquid staking tokens can look identical on a price chart and rest on opposite foundations. A decentralized LST like stETH, rETH, or frxETH lives in smart contracts. You never hand your keys to a company; the code holds the logic, and your risks are a contract bug or a governance failure. A custodial LST like cbETH or wBETH is a receipt from an exchange that stakes on your behalf. The user experience is smoother and the counterparty is a regulated business, but you are trusting that business to actually hold your coins and not rehypothecate them, a trust that Celsius and FTX taught the market not to extend lightly.

A third model matured in 2026: regulated institutional custody. When Anchorage Digital, a federally chartered crypto bank, integrated Lido in July 2026, it let institutions mint and redeem wstETH inside a compliant custody stack rather than on a public exchange. Nathan McCauley, Anchorage’s co-founder and chief executive, called liquid staking ‘one of the most important building blocks for institutional participation in Ethereum’, per news.bitcoin.com. Weeks later the Ethereum treasury company SharpLink committed to stake 200 million dollars of ETH through Lido, taking wstETH and custodying it at Anchorage, per crypto.news. The trust question does not vanish at that scale; it just moves from an exchange’s balance sheet to a bank’s.

Liquid Staking Is Not Restaking

The two terms get blurred constantly, so draw the line clearly. Staking secures Ethereum. Liquid staking wraps that staked position in a tradable token. Restaking takes staked ETH and re-pledges it to secure additional systems, oracles, bridges, data-availability layers, through a marketplace like EigenLayer or Symbiotic, in exchange for extra yield and, importantly, extra slashing risk. A liquid restaking token such as weETH is one layer higher and one notch riskier than a plain LST like stETH.

In 2026 the market delivered a verdict on that extra layer. ether.fi, once the poster child of liquid restaking, removed restaking from its flagship weETH and reverted it to a plain liquid staking token, isolating restaking in a separate opt-in product, per The Defiant. The lesson for anyone comparing yields: liquid staking is the stable base layer, and restaking is an optional bet on top of it that even its biggest champions pulled back from. If you want the full picture of that layer and why demand cooled, see our complete guide to restaking.

Liquid Staking Beyond Ethereum: The Solana Model

Ethereum is not the only chain with liquid staking, and Solana shows how different the same idea can look. Solana uses delegated proof of stake with no 32-coin minimum and no long entry queue, so around 68 percent of all SOL is already staked, one of the highest ratios of any major network, per Datawallet. Because native staking is so easy there, Solana LSTs are not selling access to staking; they are selling composability, instant liquidity, and a cut of MEV.

That is why penetration is lower. Only somewhere in the mid-teens percent of staked SOL is held in liquid form, well under Ethereum’s roughly one-third, and the leaders are different names: JitoSOL, the largest independent Solana LST, built its pitch on distributing MEV rewards to holders, while Marinade and Sanctum compete on aggregation and validator choice, per DefiLlama. One nuance carries over from Ethereum and matters for redemptions: native Solana unstaking is gated to epoch boundaries, roughly every two days, and a full cooldown can take longer, so even on a fast chain the instant liquidity of an LST is a real advantage over waiting for an epoch to turn.

The Big 2026 Shift: Validator Consolidation and Lido Core

The Pectra upgrade quietly changed the shape of Ethereum staking by raising the maximum effective balance of a single validator from 32 ETH to 2,048 ETH under EIP-7251. That lets one validator do the work that used to take dozens, compounding rewards automatically instead of sweeping them out. Through most of Ethereum’s history that option sat mostly unused. In 2026 the largest liquid staking protocol started using it at scale.

Lido’s Core 2026 overhaul migrates more than 265,000 of its validators from legacy withdrawal credentials to the new compounding format, a consolidation that shrinks Ethereum’s total validator set from roughly 880,000 toward an estimated 628,000, cutting attestation messages per epoch by about 29 percent, per cryptonews.com. The share of staked ETH secured by compounding validators is set to jump from about 32 percent to 52 percent, and the first consolidations are scheduled to begin in September 2026. The same upgrade, through a revamped Curated Module, requires Lido’s operators to post ETH bonds for the first time, replacing pure reputation with real skin in the game. The headline is efficiency, but the subtext matters more: liquid staking is now infrastructure large enough that one protocol’s housekeeping reshapes the base layer of Ethereum.

The Fight Over the Yield: EIP-8363

If liquid staking helped push the staking ratio up and the reward down, some Ethereum researchers now want to cap the ratio outright, and their tool is the yield itself. A draft proposal that surfaced on 4 August 2026, EIP-8363, nicknamed Tapered Issuance Burn, would burn a rising fraction of every validator’s consensus reward as more ETH is staked, following the formula b equals (D divided by 60,250,000) raised to the power 1.5, where D is total staked ETH. Once roughly half the supply, about 60.25 million ETH, is staked, net new issuance would reach zero. The change would phase in over 18 months, and its authors include the Ethereum Foundation’s Justin Drake and ETHCC co-founder Jerome de Tychey, per Cointelegraph.

The rationale is that additional staking past a point buys diminishing security while diluting holders who choose not to stake, and that an uncapped ratio feeds exactly the custodial and liquid-staking concentration that worries decentralization advocates. De Tychey has warned that more than 70 million ETH could be staked by January 2028 if nothing changes. The backlash was immediate and loud. Aave founder Stani Kulechov and ether.fi chief executive Mike Silagadze led a wave of opposition, arguing the burn would gut Ethereum’s lending markets, hurt solo stakers, and undermine confidence in the network’s monetary policy, per The Defiant. For now the proposal is only a draft; it has not reached the weakest formal stage of inclusion and is not part of the coming Hegota upgrade. But it points a spotlight at the exact reward that liquid staking exists to capture. If any version of it ever ships, the consensus issuance that makes up the bulk of an LST yield would shrink toward zero as staking rises, leaving tips and MEV as the only durable sources, and every LST holder should watch where the debate lands.

The Risks You Actually Take On

Liquid staking is useful, not free of danger. The risks are specific and worth naming.