Lido vs Rocket Pool vs Frax: One Giant, Two Survivors

The size gap is the whole story

Every few weeks another Lido versus Rocket Pool versus Frax scorecard makes the rounds, ranks the three on yield and fees, and crowns a winner. That framing buries the single most important fact about Ethereum liquid staking in 2026: the contest is already settled. Lido’s stETH is worth about $23.3 billion, according to CoinGecko. Rocket Pool’s rETH is worth roughly $907 million, per CoinGecko. The entire Frax Ether system holds about $317 million in total value locked. Add the two challengers together and Lido is still around nineteen times their combined size.

So the useful question is not which liquid staking token is best. It is this: one protocol already won, so what are you actually buying when you pick a smaller rival, and is Lido’s dominance the thing you should worry about most? Each of the three offers a different answer. Lido asks you to accept the incumbent and its scale. Rocket Pool sells decentralization as insurance. Frax barely competes on staking share at all; it treats staked ETH as one ingredient inside a much larger stablecoin and Layer 2 system. This piece compares them on that basis, not as three interchangeable yield tokens.

The backdrop makes the choice timely. ETH has ripped higher through late August, trading near $2,429 and up about 2.6% in a day on CoinGecko, with wstETH up roughly 29% over the week as the rally fed through to staked positions. Traders have credited a mix of expanded US Treasury buybacks (a quasi-easing some have called “QE Lite”) and renewed spot-ETF inflows, per dMarketForces. Every dollar of that move makes the question of where to park staked ETH more pressing, and it sharpens the difference between the three protocols. If you want the layer that sits on top of liquid staking, our guide to restaking in 2026 covers how that market grew and then partly unwound.

A thirty-second refresher on what these tokens do

Staking ETH means locking up 32 ETH (or a multiple of it) to run a validator that helps secure Ethereum and earns rewards for doing so. Two frictions make that awkward for most holders: the 32 ETH minimum, and a queue to get in. As of 22 August, about 42.3 million ETH is staked, roughly 34.65% of all supply, across 901,691 validators, with an entry queue holding around 2.28 million ETH and a wait near 39 days, according to ValidatorQueue. Exiting is quick again, under nine hours plus a short sweep, after the queue swung from a long backlog to near-empty over the past year. That backlog was built by ETF and corporate-treasury demand to stake, the same flows now reshaping which protocol wins, and it is exactly the friction liquid staking exists to route around.

Liquid staking removes both frictions. You deposit any amount, and the protocol hands you a token, stETH, rETH or frxETH, that represents your staked ETH plus accrued rewards and trades freely on the open market. Buying or selling that token never touches the validator queue, because you are trading an existing claim rather than spinning up or winding down a validator. The token is also composable: you can lend it, supply it to a liquidity pool, or post it as collateral, which is why liquid staking tokens became the reserve collateral of DeFi.

Two accounting models show up across the three protocols. Rebasing tokens like stETH grow in balance daily, so your wallet shows more stETH over time. Reward-bearing tokens like wstETH, rETH and sfrxETH keep a fixed balance but rise in value against ETH, which is why one rETH is worth more than one ETH. A 2025 upgrade, EIP-7251, raised the maximum effective balance per validator from 32 ETH to 2,048 ETH, letting large operators consolidate validators and compound rewards, a change that mostly benefits pools and institutions. One clarification worth keeping straight: liquid staking is not the same as restaking, which re-pledges staked ETH to secure additional services and carries its own risks.

Lido, the protocol that already won

Lido is the default. Its stETH is a rebasing token backed by around 9.5 million ETH, worth about $23.3 billion on CoinGecko, and its wrapped, non-rebasing version, wstETH, trades at 1.2440 ETH each, per CoinGecko. In practice most DeFi users hold the wrapped version, wstETH, because a rebasing balance is awkward for smart contracts, while stETH itself is the simplest hold-and-earn form; both represent the same staked ETH. Lido charges a 10% fee on staking rewards, split evenly between node operators and the DAO treasury. Rewards are spread across a curated set of professional operators, now supplemented by a permissionless Community Staking Module and Simple DVT validators that use distributed-validator middleware to reduce single-operator risk.

Lido’s share depends entirely on which denominator you use, and the two are constantly conflated. Within the liquid staking segment specifically, Lido runs somewhere between about 50% and 62% depending on the tracker, per DefiLlama and CCN. Measured against all staked ETH, the number that actually matters for Ethereum’s security debate, Lido sits closer to 23%. Both figures are true at once; quoting the higher one to imply imminent capture, or the lower one to wave the issue away, is the most common sleight of hand in this topic.

The 2026 story for Lido is a move upmarket. Lido V3, live since 30 January 2026, introduced stVaults, a modular framework that lets a staker pick its own operators, customize the validator setup, still mint stETH, and even opt out of Lido governance in the event of governance capture, as Figment details in its breakdown and Lido describes in its launch post. That sits on top of dual governance, live since 2025, which lets stETH holders veto DAO decisions and, if more than 10% of stETH is escrowed against a proposal, trigger a rage-quit that pauses governance until dissenting stakers exit, per Lido. Both features are Lido’s answer to the charge that it is simply too big.

Rocket Pool, paying a premium for decentralization

Rocket Pool sells the opposite of Lido’s curated efficiency: a permissionless validator set that anyone can join. Its rETH is a reward-bearing token worth 1.1693 ETH each, with a market value around $907 million on CoinGecko, and total protocol value locked just under $1 billion. The pitch is credible neutrality: no gatekeeper decides who runs a validator, so no single entity controls the stake.

For years the drawback was capital. A Rocket Pool node operator had to post an 8 ETH bond plus stake RPL as insurance, which limited how many people could run nodes. The Saturn 1 upgrade, live on mainnet since 18 February 2026, changed the economics. It cut the minimum bond from 8 ETH to 4 ETH and introduced megapools that let one operator run many validators under a single contract, so 8 ETH of bonded capital can now support up to 56 ETH of liquid deposits, roughly doubling capital efficiency, according to Crypto Briefing. Megapool validators also earn about 2.3 times more commission per bonded ETH, and, crucially, Saturn switched on protocol revenue for the RPL token, as The Defiant explains.

Markets noticed. RPL trades around $1.67 with a market cap near $38 million on CoinGecko, up more than 20% on the week as the rally and the fee switch re-rated the token. The tradeoff for choosing rETH over stETH is real: less liquidity, fewer DeFi integrations, and historically a slightly lower net yield after node-operator commission, in exchange for a validator set no one can quietly capture. Rocket Pool’s own documentation is the cleanest primary source on how the bond and commission model works. If you think of proof-of-stake as a security budget the way proof-of-work miners do, our look at hashprice and Bitcoin’s security bill is the mining-side mirror of the same question: who secures the chain, and who pays for it.

Frax, quietly opting out of the popularity contest

Frax is the odd one out, and deliberately so. Instead of one token, it runs two. frxETH is a peg token designed to replace wrapped ETH inside smart contracts and liquidity pools; it pays no staking yield on its own. sfrxETH is an ERC-4626 vault that collects the staking rewards of the entire system and concentrates them into the minority who actually stake, so sfrxETH’s yield is structurally boosted by every frxETH holder who chooses liquidity over yield. The whole Frax Ether system holds about $317 million in total value locked, with sfrxETH trading at 1.1529 ETH each on CoinGecko. The design is documented in the Frax docs, and frxETH and sfrxETH are wired into Curve, Convex, Yearn and Pendle.

The bigger shift came at the end of 2025. Frax’s North Star upgrade, announced 30 December 2025, renamed the governance token from FXS to FRAX and repurposed it as the native gas token of Fraxtal, the project’s Layer 2, with holders on Ethereum bridging to convert, per PANews. Founder Sam Kazemian has framed the pivot as reflecting what Frax has become, a stablecoin operating system rather than a single protocol, and he has been visibly involved in US stablecoin legislation around the GENIUS Act.

Read that way, Frax is not really trying to beat Lido on staking share. frxETH is one lego brick inside a system that also includes the frxUSD stablecoin (backed in part by BlackRock’s BUIDL fund), the Fraxtal L2, and a permissionless lending market in which node operators borrow ETH at a variable rate to run validators, so sfrxETH earns staking yield and lending interest at once. If you want a straightforward stETH substitute you would not start here; if you want an ETH-denominated building block that plugs into a broader DeFi and stablecoin machine, or you want sfrxETH’s boosted yield, Frax makes more sense than its size suggests.

The scoreboard

Here is the head-to-head, with live figures from CoinGecko as of 22 August 2026. Note that Lido’s size is stETH market value, Rocket Pool’s is rETH market value, and Frax’s is whole-system TVL, so the columns are close proxies rather than identical measures.

Where the yield actually comes from

Start with the base. Ethereum’s staking APR has fallen to roughly 2.6% to 2.7% on the consensus layer, down from a 5.06% peak in June 2023, with execution-layer tips and MEV lifting well-run validators to something like 3% to 3.8%, per CoinPedia. That decline is simple math: the more ETH staked, the thinner the rewards are spread, and staking just hit an all-time-high share of supply. Every liquid staking token starts from that same base and then differs in how it packages it.

Lido takes the base yield, skims its 10% fee, and spreads MEV across the whole pool, so stETH tracks the network rate closely; DefiLlama pegs stETH’s supply APY around 2.2%. Rocket Pool passes the base yield to rETH holders after node operators take their commission, and post-Saturn a slice of protocol revenue now flows to RPL stakers as well. Frax does something structurally different: because non-staking frxETH holders forfeit their share, sfrxETH concentrates the system’s yield into fewer tokens and can screen higher than the naive base rate, before adding Curve and AMO strategy revenue on top.

The question to ask of any staking yield is whether it is real or subsidized. Real yield comes from fees and MEV, meaning someone actually paid for a service; subsidized yield comes from token emissions that dilute you. Ethereum liquid staking yields are overwhelmingly real, funded by consensus and execution rewards, which is a genuine advantage over some Solana liquid staking tokens that lean on inflationary subsidies. Leverage can amplify any of these through looping, but it adds liquidation risk that turns a quiet 3% into a fast way to lose your principal in a depeg.

The one-third problem is no longer theoretical

Ethereum has soft thresholds that make concentration dangerous. An entity controlling more than one third of staked ETH can threaten the chain’s ability to finalize; one half enables censorship; two thirds could finalize an invalid chain. The clearest early write-up of this risk came from Ethereum Foundation researcher Danny Ryan in a widely cited note on the risks of liquid staking. With Lido near 23% of all staked ETH, the first threshold is no longer a whiteboard exercise; it is a number the community watches every month.

Vitalik Buterin has repeatedly named staking centralization as one of the biggest risks to Ethereum as part of his Scourge research track, warning that a single dominant liquid staking provider could capture the network’s economic gravity, as The Block reported. That is the strategic case for the challengers existing at all: a world with several mid-sized liquid staking tokens is healthier for Ethereum than a world with one giant, even if the giant is well run.

Lido’s response has been to build guardrails rather than cap its size. It declined to self-limit when the question went to a DAO vote, and instead shipped dual governance so that stETH holders can veto or rage-quit, and V3 stVaults so that stakers can opt out of Lido governance entirely. Critics counter that neither feature actually reduces Lido’s raw share of the validator set, which is the number the thresholds care about. That unresolved tension is exactly why Rocket Pool’s permissionless model reads as insurance, and why picking a smaller token can be a vote for Ethereum’s health even when it costs you a little yield or liquidity.

The upgrades that reset the math in 2026

All three protocols shipped a flagship upgrade inside twelve months, and the striking thing is that they point in three different directions. Lido moved upmarket toward institutions, Rocket Pool doubled down on scaling decentralization and finally rewarded its token, and Frax reframed itself as a stablecoin and Layer 2 system that happens to include a staking token. Read together, the upgrades confirm the three-strategies thesis better than any marketing deck could.

The risks nobody prices until they bite

Liquid staking tokens are not risk-free versions of ETH, and the failure modes are specific. The first is a depeg. In May 2022, as Terra collapsed, Three Arrows Capital pulled roughly 128,000 stETH and 73,000 ETH out of the main Curve pool in a single transaction, thinning exit liquidity and sending stETH from about 97 cents on the dollar to the low 90s, with no hack involved, purely a liquidity and confidence event, as CoinDesk reported at the time. Crucially, stETH was not yet redeemable then; today all three tokens support in-protocol withdrawals, which anchors the peg far better than in 2022.

The second is oracle and contagion risk, which shows up once these tokens become collateral. In March 2026, a stale price-oracle timestamp on Aave undervalued wstETH by about 2.85%, triggering roughly $27 million in otherwise avoidable liquidations across 34 accounts, with zero bad debt and a full reimbursement pledge from the risk manager, per The Block. There was no real depeg; the oracle, not the token, was the point of failure, a reminder that leverage built on liquid staking tokens can break even when the token itself is fine.

The rest of the risk map is structural. All three protocols are smart contracts, exposed to bugs, upgrade-key compromise and validator slashing; if you want a sense of how often audited code still gets hit, our profile of Halborn is a sober read. On top of that, Lido carries concentration and governance-capture risk, Rocket Pool carries liquidity and RPL-collateral complexity, and Frax carries dependency risk from its many interlocking parts and its exposure to the wider stablecoin system. The table below is a qualitative map, not a scoreboard; every entry is a tradeoff, not a verdict.

Can you actually get your ETH back?

The whole point of a liquid staking token is that you can leave without joining the validator exit queue, but leaving has two doors. You can redeem through the protocol, burning the token for ETH, which is subject to that exit queue; or you can sell on the open market, which is instant but pays whatever price buyers are offering. Since the Shapella upgrade in April 2023, all three protocols support in-protocol redemption, and that is what keeps the market price anchored: if a token trades below its redemption value, arbitrageurs buy it, redeem it, and pocket the difference, closing the gap. The exit queue is short at the moment, under nine hours plus a sweep, so redemption is fast today; in a genuine rush for the exits, though, the queue lengthens and the tradable market price is what you actually get.

This is where the size gap bites again. stETH has by far the deepest secondary market, so a large holder can sell meaningful size with limited slippage, and wstETH is the form most lending venues price against. rETH and frxETH trade in thinner pools, so a large exit moves the price more, which is the concrete meaning of the liquidity risk the earlier table labels in the abstract. For a retail-sized position it rarely matters; for a treasury unwinding tens of millions of dollars, this difference between the three is often the one that matters most, and it is another reason institutional money keeps pooling into Lido.

Wall Street shows up

The most important development of 2026 is not a yield tweak; it is who started staking. On 2 July 2026, Anchorage Digital, a federally chartered crypto bank, integrated Lido so institutional clients can mint and burn wstETH without leaving regulated custody. Anchorage co-founder and chief executive Nathan McCauley said that “liquid staking has become one of the most important building blocks for institutional participation in Ethereum,” per Bitcoin.com News. Weeks later, on 13 August 2026, the ETH treasury company SharpLink said it would stake $200 million of ETH through Lido and hold the resulting wstETH; co-chief executive Joseph Chalom, formerly of BlackRock, called it “an exciting expansion in making our ETH even more productive, leveraging wstETH’s composability while maintaining institutional-grade risk standards,” according to Lido.

The fund wrappers arrived too. BlackRock launched its iShares Staked Ethereum Trust, ticker ETHB, on 12 March 2026 with staking baked in from inception, as CoinDesk reported. Almost all of this institutional flow favors Lido, because institutions want the deepest liquidity, the most integrations and a token their custodian already supports, which widens the gap rather than narrowing it. The other institutional route, custodial staking through an exchange, is a genuine alternative for some; we put the major venues head-to-head in our Coinbase versus Binance versus Kraken versus OKX staking yield test. The pattern to watch is that each of these deals routes through wstETH specifically, the composable, non-rebasing token a fund administrator can hold, price and lend against without wrestling a daily-changing balance, which is a quiet but real moat for Lido over rivals whose wrapped tokens carry less integration.

What the SEC actually said

US policy flipped from hostile to permissive in a little over two years. The low point was February 2023, when the SEC fined Kraken $30 million and shut down its US staking-as-a-service program, a settlement that chilled the entire market, per the SEC. The turn came in 2025. In May, the Division of Corporation Finance said protocol staking, whether solo, delegated or custodial, is not itself a securities transaction, and Commissioner Hester Peirce backed it with a companion statement titled “Providing Security is not a ‘Security’,” per the SEC.

The decisive one for this topic came on 5 August 2025, when Corp Fin extended that reasoning explicitly to liquid staking and to liquid staking receipt tokens such as stETH, per the SEC. The guardrails matter: a provider that guarantees a fixed return, or exercises discretion over when and how much to stake, can fall right back outside the safe harbor. That distinction is why BlackRock and Anchorage could move when they did. In Europe the framing is different, since MiCA regulates crypto-asset service providers and issuers rather than the staking protocol itself, leaving decentralized liquid staking in a grey zone; our report on DeFi compliance in 2026 traces where regulators are trying to attach responsibility.

So which one should you hold?

There is no single right answer, only a match between your priorities and each protocol’s tradeoffs. If you want maximum liquidity, the widest DeFi integrations and a token your custodian supports, Lido is the path of least resistance, at the cost of adding to the concentration everyone is worried about. If you value censorship-resistance and a validator set no one can capture, and you accept smaller size and slightly lower net yield, Rocket Pool is the principled pick. If you want an ETH building block that lives inside a broader stablecoin and Layer 2 system, or you specifically want sfrxETH’s boosted yield, Frax fits despite its size. And if you would rather not touch DeFi at all, custodial staking through a regulated exchange is a legitimate fourth option, with counterparty and regulatory tradeoffs of its own.

One point cuts across all of these. The choice that is individually most convenient, Lido, is also the one that makes Ethereum’s concentration problem worse, while spreading stake across smaller tokens is collectively healthier but individually a little less convenient. That is a rare and honest tension, and it is worth weighing alongside the yield and fee numbers rather than after them.

A practical middle path is to split. Nothing stops you holding stETH for the deep liquidity you want in DeFi and rETH for the portion you would rather keep on a censorship-resistant base, and that split does double duty by nudging the network away from single-provider dominance. Whichever you pick, size the position so you can survive a temporary depeg without being forced to sell at the bottom, watch the venues where you have posted the token as collateral, and treat the yield as payment for real risks rather than a free coupon.

Frequently Asked Questions

Is stETH, rETH, or frxETH safer?

None is risk-free, and they fail in different ways. stETH has by far the deepest liquidity and the most redemption depth, which helps it hold its peg under stress. rETH sits on the most decentralized operator set, so it carries the least single-entity risk. frxETH depends on the health of the wider Frax system. All three share smart-contract, slashing and depeg risk, and none removes ETH’s own price volatility.

Which has the highest yield, Lido, Rocket Pool, or Frax?

Base Ethereum staking pays roughly 2.6% to 3.8% once execution rewards and MEV are included, and after fees the liquid staking tokens land in the low-to-mid single digits. Frax’s sfrxETH can screen higher because non-staking frxETH holders forfeit their yield to it, and leverage or looping can raise the headline number while adding liquidation risk. Small differences in quoted APY are usually less important than liquidity and risk.

Why is Lido so much bigger than Rocket Pool and Frax?

First-mover advantage, the deepest liquidity, the most DeFi integrations and, increasingly, institutional custody rails all compound into a network effect. Lido’s stETH is worth about nineteen times Rocket Pool’s rETH and the Frax Ether system combined, and institutional flows in 2026 have widened rather than narrowed that gap.

Is liquid staking legal in the US?

As of August 2025, the SEC’s Division of Corporation Finance stated that protocol staking and liquid staking, including receipt tokens like stETH, are generally not securities transactions, provided the provider does not guarantee returns or exercise discretion over staking. That is a staff position, not a law, and it is not legal advice; specifics depend on how a given product is structured.

Can I lose money staking with these protocols?

Yes. You can lose money through a token depeg, validator slashing, an oracle or contagion event when the token is used as collateral, a smart-contract bug, or simply a fall in the price of ETH itself. Liquid staking reduces some frictions, but it does not turn staked ETH into a risk-free asset.

By Yuki Tanaka, senior markets writer at HOGE Wire.