Lido vs Rocket Pool vs Frax: Three Bets on Staked ETH

Ethereum just had its loudest week of the summer, and it dragged a quiet corner of the market back into view: liquid staking. This is a field guide to the three protocols that dominate it on Ethereum, Lido, Rocket Pool and Frax, and to the reason a straight side-by-side keeps confusing people in 2026. These are no longer three versions of the same product. They have become three different answers to one question: what is staked ETH actually for?

A 22% week put staked ETH back in the spotlight

ETH climbed from under $1,900 to about $2,288, a gain of roughly 22% in seven days and its strongest stretch since spring. The catalyst was macro, not crypto-native: traders piled back in on expectations of looser US dollar liquidity, helped by a two-day run of spot ETF inflows, and a large wave of short liquidations poured fuel on the move. One session alone saw ETH jump about 15%. If you want the plumbing behind that kind of week, our look at how crypto trades the Warsh Fed maps the liquidity story in depth.

A rally like that matters here for a simple reason. Around 42.2 million ETH, about 34.6% of the entire supply, is now locked in validators, and a large slice of it is held not as raw ETH but as a liquid staking token: a receipt that keeps earning staking rewards while staying tradable and usable across DeFi. When ETH jumps 22%, every one of those receipts is suddenly worth 22% more in dollars, and the question of which receipt to hold gets loud again.

The three names that own that conversation are Lido, Rocket Pool and Frax. The rest of this guide explains how each one works, how big it actually is, what you keep after fees, where the risks hide, and why the right choice depends less on yield than on which of the three bets you want to make.

Liquid staking, in one paragraph

Regular Ethereum staking asks a lot. You need 32 ETH for a full validator, your own node running around the clock, and a spot in an entry queue that currently runs about 38 days. Miss maintenance and penalties nibble at your balance; get slashed for a serious fault and the loss is far worse. Liquid staking removes almost all of that friction. You deposit ETH into a protocol, it runs professional validators on your behalf, and it hands you a token that represents your deposit plus the rewards piling up behind it. You keep the yield, you keep your liquidity because the token can be sold or lent at any time, and you skip the validator queue entirely, since trading the token happens on the open market and never touches the deposit or exit lines.

That single trick, turning an illiquid two-month commitment into a same-day-liquid asset, is why liquid staking tokens have absorbed a large share of all staked ETH. It is the proof-of-stake cousin of the yield math that Bitcoin miners live and die by; if you have read our breakdown of what the difficulty ribbon says about mining, this is the same fight for native yield, just with validators and slashing instead of ASICs and electricity. The tradeoff is trust: you are relying on a protocol’s smart contracts and its operators instead of your own hardware.

The scoreboard: three very different sizes

Before any philosophy, the raw numbers, because the size gap is the most important single fact in this comparison.

Read that table twice. Lido is not merely the leader; it is bigger than every rival combined. Its stETH alone is worth more than twenty times Rocket Pool’s rETH and roughly seventy times the entire value locked in Frax Ether. By the standard segment measure, Lido is about half of the whole liquid staking token market by itself (DefiLlama), while sitting near a quarter of all staked ETH on the network (Datawallet). So this is not a three-horse race for the same prize. It is one giant and two very different specialists, and the specialists are interesting precisely because they are not trying to beat Lido at its own game.

That gap is not just bragging rights; it changes the user experience. Depth of liquidity determines how much you can buy or sell without moving the price, how tight the spread is when you exit, and how many lending markets, exchanges and layer-2 networks accept the token as collateral. stETH and wstETH are integrated almost everywhere in DeFi, which is a real, compounding advantage: the more places a token is accepted, the more useful it is, and the more useful it is, the more places accept it. Rocket Pool and Frax have to win on something other than ubiquity, which is exactly why they have leaned into decentralization and composability respectively.

Lido’s bet: liquid staking as institutional infrastructure

Lido’s wager is that liquid staking grows up into regulated, institution-grade infrastructure, and that the winner is whoever institutions can actually use. Everything about the protocol in 2026 points that way. When you deposit ETH you receive stETH, a rebasing token whose balance grows a little every day as rewards land, and most DeFi users then wrap it into wstETH (1 wstETH is about 1.2441 ETH), a version whose quantity stays fixed while its exchange rate climbs. Behind the token, Lido runs a curated set of roughly thirty professional node operators, supplemented since 2024 by a Community Staking Module and Simple DVT clusters that let smaller and distributed operators in. The protocol takes a 10% cut of staking rewards, split evenly between those operators and the DAO treasury.

The 2026 upgrades sharpened the institutional pitch. Lido V3, live since 30 January 2026, introduced stVaults: segregated, customizable staking vaults with dedicated validators, which let a fund or a custodian stake under its own risk, compliance and reporting rules rather than pooling anonymously with everyone else. That modular design is what unlocked the year’s headline deals.

In July, Anchorage Digital, a federally chartered US crypto bank, integrated Lido so institutions could mint and redeem wstETH inside a regulated custodian. Anchorage co-founder and CEO Nathan McCauley called liquid staking “one of the most important building blocks for institutional participation in Ethereum” (News.bitcoin.com). Weeks later, the Nasdaq-listed treasury company SharpLink committed to stake roughly $200 million of ETH through Lido and hold wstETH, custodied at Anchorage. Co-CEO Joseph Chalom, a former BlackRock executive, described the point as “making our ETH even more productive, leveraging wstETH’s composability while maintaining institutional-grade risk standards” (Lido blog). This is the same current pulling traditional finance on-chain that we traced in our report on RWA lending, only here the asset being plugged in is staked ETH itself.

Rocket Pool’s bet: a permissionless public good

Rocket Pool is making the opposite bet: that the value of liquid staking is trust minimization, and that the network needs a version anyone can help run without permission. Where Lido curates its operators, Rocket Pool lets anyone become one. Hold enough ETH and its RPL collateral token, spin up a node, and you are validating. Depositors who just want yield receive rETH, a reward-bearing token that holds a steadily rising amount of ETH (1 rETH is about 1.1685 ETH) with no rebasing and no active management required.

The protocol’s big 2026 event was the Saturn I upgrade, live 18 February 2026. It cut the minimum node operator bond from 8 ETH to just 4 ETH per validator, so a node runner now supplies 4 ETH and the pooled rETH side supplies the other 28 to form a 32 ETH validator. It also introduced megapools, which consolidate many validators under one contract to save gas, an express queue for experienced operators, and a new commission scheme called the Universal Adjustable Revenue Split (Saturn). Under it, operators earn a base commission of roughly 5% on the pooled stake plus an additional share, up to around 9% more, distributed in proportion to how much RPL they stake, and RPL is no longer mandatory just to launch a validator.

The result is a protocol that is far smaller than Lido, around 1% of all staked ETH, but structurally the most decentralized of the three: thousands of independent operators rather than a curated few, and a token, rETH, that is a claim on that distributed validator set rather than on a single company’s roster. The governance token RPL trades near $1.55, a reminder that decentralization has not translated into a rich token, a tension Rocket Pool shares with plenty of infrastructure projects.

The RPL token sits at the center of Rocket Pool’s design and its ongoing debate. Historically, operators had to stake RPL worth at least 10% of their borrowed ETH as an insurance buffer against penalties, which tied demand for the token to the growth of the network. Saturn loosened that link by making RPL optional to launch a validator while still rewarding operators who stake it with a larger slice of commission. Supporters say separating validation from token speculation is healthier; skeptics note it also removes a structural bid for RPL, which helps explain why the token has languished while rETH has grown. It is the classic infrastructure tension: the network can succeed even if the governance token does not capture much of that success.

Frax’s bet: staked ETH as a money lego

Frax is not really trying to win the liquid staking market at all. For Frax, staked ETH is one brick inside a larger structure that also includes a stablecoin, an automated market operations engine and its own layer-2 network. That is why its design looks stranger than the others. Frax splits the job across two tokens. frxETH is a pure peg token that tracks ETH roughly one to one (about $2,281 against ETH’s $2,287) and by itself earns no staking yield; it is meant to be used as liquidity and collateral. sfrxETH is the yield vault: stake your frxETH into it and you receive an ERC-4626 vault token whose value climbs faster than a normal LST, because all the staking rewards that the many frxETH holders forgo get concentrated into the smaller pool of sfrxETH holders (1 sfrxETH is about 1.1588 ETH).

The frxETH v2 design goes further, turning validation into a permissionless lending market: node operators borrow ETH at a variable rate to run validators, sfrxETH captures the staking yield plus that lending interest, and Curve liquidity is managed by the protocol’s own algorithms. In late 2025 Frax pushed through its North Star hardfork, which retired the old FXS ticker and rebranded the governance token to FRAX, now the gas token of the Frax-built Fraxtal layer-2; major exchanges finished migrating the ticker by January 2026 (CoinGecko). Frax also runs frxUSD, a stablecoin it has backed with tokenized assets including BlackRock’s BUIDL fund. Founder Sam Kazemian has been consistent that the goal is a self-contained on-chain dollar-and-yield system, with frxETH as its ETH-denominated engine rather than a standalone product. Frax Ether’s total value locked, about $310 million, is the smallest of the three by a wide margin, which is the point: it is a component, not a destination.

To see why Frax builds this way, look at the rest of the stack. Fraxtal, its OP-Stack layer-2 that went live in early 2024, uses FRAX for gas and is designed to route the fees and yield it generates back into the Frax ecosystem. The protocol’s automated market operations, or AMO, contracts deploy idle collateral into strategies like Curve liquidity and lending, so a single deposit can be productive in several places at once. In that context frxETH is less a competitor to stETH than a yield-bearing reserve asset for a self-contained on-chain economy. That makes Frax the hardest of the three to evaluate as a pure staking choice, because you are really buying into a whole financial system, not just a validator set.

Rebasing vs reward-bearing: the design that trips people up

The most common source of confusion in this whole category is not yield or fees; it is how the tokens report your rewards. There are two schools. Rebasing tokens like stETH keep a price close to 1 ETH and grow your balance instead: hold 10 stETH today and you might see 10.02 next month, with the extra representing rewards. Reward-bearing tokens like wstETH, rETH and sfrxETH keep your balance fixed and let the token’s exchange rate rise, so one token is worth more ETH over time. Frax’s frxETH is a third case, a peg token that deliberately earns nothing on its own.

The distinction is not cosmetic. Rebasing balances are awkward inside many DeFi contracts and can complicate accounting, which is why serious DeFi users and most institutions prefer the reward-bearing wrappers. It also matters at tax time: in the US, staking rewards are generally taxable as income when you gain control of them, and a rebasing token that credits you daily creates a very different paper trail from one that simply appreciates. With brokers now reporting to the IRS, that distinction is no longer academic; our guide to crypto tax in the 1099-DA era walks through why the form of your rewards can change what you owe.

Yield, fees, and what you actually keep

The uncomfortable truth of 2026 staking is that the raw yield is not exciting. Base consensus rewards have drifted down to roughly 2.6% to 2.8% a year, well off the 5% highs of 2023, because so much ETH is now staked that the per-validator reward has thinned. On top of that base, validators earn priority fees and MEV, which add a little more, and then each protocol takes its cut.

Two things follow. First, the fee model shapes the net yield more than the headline APR does: Lido’s flat 10% is simple and predictable, Rocket Pool’s split rewards operators who stake RPL, and Frax’s structure deliberately routes yield away from frxETH and into sfrxETH. Second, and more important for a dollar investor, that 2.6% base now sits below the US risk-free rate; short-term Treasuries have been yielding more than ETH staking for months. The honest case for staking in 2026 is not the yield in isolation; it is the yield on top of an asset you already want to hold for price upside, which the past week’s 22% move makes vivid. If your thesis is purely income, the macro backdrop matters as much as the protocol you pick.

The peg: what 1 stETH = 1 ETH really means

Newcomers assume a liquid staking token is always redeemable for exactly one ETH. It is not, and the gap between the exchange rate and the market price is where the real risk lives. Each token has an underlying, protocol-defined exchange rate: one stETH is backed by one ETH of stake, one rETH by about 1.17 ETH. But the market price can wander from that backing when everyone wants out at once and liquidity is thin. The textbook example is May 2022. As the Terra collapse spread panic, Three Arrows Capital pulled roughly 128,000 stETH out of the main Curve pool in a single move, thinning the exit door, and Celsius was sitting on a large stETH position it could not unwind. stETH slid from about 97 cents on the ETH to the low 90s within days (Nansen via CoinDesk).

Crucially, that was a liquidity and confidence event, not an insolvency: every stETH was still fully backed, but at the time stETH was not yet redeemable, so holders who needed cash immediately had to sell into a shallow market. That structural weakness is largely gone now. Ethereum’s Shapella upgrade enabled staking withdrawals in April 2023, so all three protocols support redemption, and the near-empty exit queue today (well under an hour to leave) means a discount can be arbitraged away quickly. The lesson still stands, though: a liquid staking token can trade at a discount in exactly the moments you most want to sell, and that discount is a feature of market microstructure, not a sign the protocol is broken.

In practice, there are two numbers worth watching for any of these tokens. The first is the protocol exchange rate, the amount of ETH each token is redeemable for, which only ever climbs as rewards accrue. The second is the live market price on exchanges, which can sit slightly above or below that rate depending on demand and liquidity. A small premium usually means buyers want yield exposure quickly; a discount means sellers are willing to pay for an instant exit rather than wait in the redemption queue. Comparing the two tells you whether you are paying up for convenience or being paid to provide it, and it is the single most useful habit a liquid staking holder can build.

Composability: where these tokens actually go

The reason liquid staking tokens matter beyond convenience is composability: they have become base collateral across DeFi. wstETH in particular is one of the most-deposited assets on lending markets like Aave, where users borrow against it, and where some run the leveraged strategy known as looping (deposit wstETH, borrow ETH, buy more wstETH, repeat) to amplify a 2.6% base into high-single-digit or low-double-digit returns, at the cost of liquidation risk if the token slips. That machinery is powerful and fragile in equal measure. In March 2026, a safeguard oracle that Aave uses for wstETH briefly reported a stale value, undervaluing the token by about 2.85% and triggering roughly $26 million in avoidable liquidations across 34 accounts before it was corrected; there was no bad debt, and Chaos Labs pledged to reimburse affected users (The Block).

That incident is the cleanest reminder that the risks here are rarely about the staking itself. The staking worked fine; an oracle and a liquidation engine did not. Most catastrophic losses in this space come not from broken cryptography but from the plumbing around it, and above all from compromised keys and access, the theme of our piece on why stolen keys beat broken code. When you hold a liquid staking token that is levered three times over inside a lending market, you inherit the smart-contract risk of the protocol, the oracle, the lending market and any bridge in between, not just Lido or Rocket Pool or Frax.

Composability also reaches beyond lending. Protocols like Pendle let holders split a liquid staking token into its principal and its future yield, then trade each separately, so a staker can lock in a fixed rate while a speculator bets on rewards rising or falling. Others use wstETH as the reserve backing for stablecoins and structured products. Each layer adds utility and a new dependency, and the further a token travels from the validator that mints it, the more its price reflects conditions in those downstream markets rather than the staking yield itself. For most holders the practical takeaway is simple: know exactly how many protocols sit between you and the underlying ETH, because every one of them is a place something can break.

The one-third problem: Lido’s size is a network question

Lido’s dominance is not only a competitive fact; it is a live governance debate for Ethereum itself. Ethereum’s consensus has meaningful thresholds. A single entity controlling more than one-third of all staked ETH can, in the wrong circumstances, prevent the chain from finalizing; past one-half it gains censorship power, and past two-thirds it could finalize an invalid chain. Lido sits near a quarter of all staked ETH today, below the one-third line, but close enough that researchers keep raising it. Vitalik Buterin has repeatedly named staking centralization “one of the biggest risks” to Ethereum as part of his Scourge research track (The Block).

Lido’s answers are twofold. First, it distributes stake across many independent operators and DVT clusters, so it is not one validator but a coordinated set, which softens part of the concern. Second, in July 2025 it activated a dual-governance system that gives stETH holders a way to veto or delay DAO decisions they oppose, adding a check on the token holders who run governance. Critics counter that operator diversity does not fully address the protocol-level concentration, and that the market, not a single DAO, should keep any one provider under the line. It is a genuine open question, and it is the strongest structural argument for the existence of a permissionless alternative like Rocket Pool, even at a fraction of the size.

This is not a new argument, and Lido has faced it directly before. Back in 2022 the Lido DAO put a self-limiting proposal to a vote, asking whether the protocol should cap its own share of staking to protect the network; holders overwhelmingly rejected the cap, arguing that a voluntary limit would simply hand share to less transparent competitors. Whatever you make of that reasoning, it framed the debate that still runs today: should the market discipline a dominant provider, or should the provider restrain itself? Rocket Pool’s permissionless model and Frax’s smaller footprint are, in effect, two different answers to the same worry, which is why the health of the alternatives matters even to people who only hold stETH.

Regulation: the SEC quietly cleared the path

For years the biggest cloud over US staking was regulatory. That cloud has largely lifted. In May 2025 the SEC’s Division of Corporation Finance stated that protocol staking, whether solo, delegated or custodial, is not itself a securities transaction, and Commissioner Hester Peirce backed it with a statement pointedly titled “Providing Security is not a ‘Security’” (SEC). A follow-up in August 2025 went further and extended that view to liquid staking and the receipt tokens it produces, exactly the stETH-and-rETH category this guide is about (SEC). The guardrails are narrow but clear: a custodian that guarantees a fixed return, or that exercises discretion over when and how much to stake, can fall back outside the safe harbor.

That legal clarity is what made 2026’s institutional wave possible. BlackRock launched a staked-ether ETF, trading as ETHB, in March 2026, with staking baked in from day one, and the Anchorage and SharpLink deals followed the same logic. None of that clears the tax question, though. Staking rewards remain taxable income when received, and the mechanics differ by token type, a distinction with a direct cost at filing time now that broker reporting is automatic.

The risk that could reprice all three: EIP-8361

The largest cloud on the horizon is not regulatory or technical; it is monetary policy, Ethereum’s own. A proposal known as EIP-8361, drafted in August 2026 by a group of researchers, would taper validator issuance toward zero as the staking ratio rises, reaching no net issuance at around 60.25 million ETH staked, roughly half the supply. The goal is to put a natural cap on how much ETH gets staked, on the theory that runaway staking pushes ever more ETH into a handful of custodial and liquid staking providers, the very concentration the previous section described.

For liquid staking, the implication is direct: if issuance falls, so does the base yield that stETH, rETH and sfrxETH pay, and the leveraged strategies built on top of them get squeezed. Aave founder Stani Kulechov warned that pushing rewards toward zero would make ETH borrowing strategies “mostly unviable,” while other developers argued it would push out exactly the solo stakers Ethereum wants to keep. The proposal did not make the near-term upgrade cut and would phase in over roughly two years if adopted, so nothing is imminent. But every serious staker should understand that the yield underpinning all three protocols is a policy variable the community is actively debating, not a fixed constant.

How to choose: matching the protocol to the person

Because the three protocols are making different bets, the right pick depends on who you are, not on a single leaderboard.

If you want the deepest liquidity, the widest DeFi acceptance and the smoothest institutional on-ramps, Lido is the default, and wstETH is the form to hold. If your priority is decentralization and censorship resistance, and you are comfortable with thinner liquidity, Rocket Pool’s rETH is the most trust-minimized token in the category, and Saturn’s 4 ETH bond makes actually running a node realistic for more people. If you live inside DeFi, build with stablecoins, or want the most concentrated yield and do not mind added complexity, Frax’s sfrxETH is the specialist’s tool. And if you are a treasury or a fund, Lido V3 stVaults plus a qualified custodian is the combination that most closely matches how you are already required to operate. None of these is the wrong answer; they are answers to different questions.

Frequently Asked Questions

Is Lido or Rocket Pool better for staking ETH?

Neither is universally better; they optimize for different things. Lido is far larger, more liquid and more widely accepted across DeFi and by institutions, so it is the practical default for most holders. Rocket Pool is smaller but more decentralized, with thousands of permissionless node operators, so it appeals to users who prioritize censorship resistance and trust minimization over liquidity. If you want convenience and integrations, choose Lido; if you want decentralization, choose Rocket Pool.

What is the difference between stETH, rETH and frxETH?

stETH is Lido’s rebasing token: its balance grows daily and it stays close to 1 ETH. rETH is Rocket Pool’s reward-bearing token: its balance is fixed and its value rises, so one rETH is now worth about 1.17 ETH. frxETH is Frax’s peg token that tracks ETH one to one and earns no yield by itself; to earn, you stake it into sfrxETH, a vault that concentrates the rewards other frxETH holders give up.

Is liquid staking safe?

The staking itself is well understood, but liquid staking adds layers of risk: smart-contract bugs, operator failures that can trigger slashing, and market depegs when everyone sells at once, as stETH did in 2022. The larger dangers usually come from how the tokens are used, for example leveraged looping on lending markets, or from compromised keys and faulty oracles rather than the staking. Redemption is now enabled and exit queues are short, which reduces depeg risk, but liquid staking is not risk-free.

Do you pay tax on stETH and other liquid staking rewards?

In the US, staking rewards are generally treated as ordinary income at the time you gain control of them, and later disposals can trigger capital gains. Rebasing tokens like stETH credit rewards continuously, while reward-bearing tokens like wstETH and rETH appreciate in value, and the two can create different reporting trails. With brokers now issuing 1099-DA forms, this is reported automatically, so keeping records of cost basis matters more than ever.

Can Lido become too big for Ethereum?

It is a real concern. A single entity controlling more than one-third of all staked ETH could, in bad conditions, interfere with the chain finalizing, and more than one-half or two-thirds carries worse risks. Lido sits near a quarter of all staked ETH, below that line, and spreads stake across many operators and DVT clusters, plus a dual-governance veto for stETH holders. Critics argue operator diversity does not fully solve protocol-level concentration, which is the strongest argument for keeping decentralized alternatives like Rocket Pool healthy.

Yuki Tanaka covers staking, validators and DeFi infrastructure for HOGE Wire.