Liquid Staking Tokens (LSTs): A Research Guide

State of liquid staking in 2026

35.86M ETH (28.91% of supply) is staked across 1.1M+ validators; Lido controls ~24% of all staked ETH, down from a 32% peak in 2023. Base consensus yield is 2.84%, per the Compass STYETH index (May 1, 2026). Last verified: 2026-05-27.

35.86M ETH is staked — 28.91% of total supply — secured by 1.1M+ validators, per Datawallet. Lido (Lido is the largest liquid staking protocol on Ethereum, issuer of stETH) now controls ~24% of all staked ETH (~8.7M ETH) — down from a 32% peak in 2023 and a ~28% figure cited in early-2026 sources, as Ether.fi, Binance, and centralised-exchange staking collectively gained share. The total LST (LST = liquid staking token: a tokenised receipt for ETH staked through a protocol that accrues yield while remaining usable in DeFi) market cap exceeds $30B.

Base ETH consensus-layer yield is 2.84% as of May 1, 2026 (Compass STYETH index) — compressed steadily from ~4.5% at the post-merge peak as the validator set grew to 1.1M+. MEV (MEV = maximal extractable value, the additional revenue validators earn from transaction ordering) and priority fees add roughly 0.3–0.5% on top; the exact realised boost depends on block demand. Lido stETH and Rocket Pool rETH have multi-year track records with no protocol-level losses; newer LSTs and LRTs (LRT = liquid restaking token, an LST further restaked via EigenLayer for additional rewards) compete on yield differentiation but carry materially more risk.

For most ETH holders, plain stETH or rETH beats every LRT, points-farm, and exotic-LST on risk-adjusted yield.

Top 5 LSTs in 2026

Lido stETH (~24% of staked ETH, deepest DeFi liquidity) leads; Ether.fi weETH ($7–9B TVL) is now #2 by size; Rocket Pool rETH (3,200+ operators, Saturn One live) is the decentralisation pick. Last verified: 2026-05-27.

TVL figures sourced from DefiLlama and vaults.fyi as of late May 2026; ranges reflect intra-month variance. Ether.fi weETH is included here for context — it is an LRT (liquid restaking token), not a plain LST.

Lido (stETH) — deep dive

Full guide: Lido.

stETH has universal DeFi acceptance — every major protocol is built to handle it. That liquidity moat is also the strongest argument against the smaller LSTs.

Best for

DeFi power users who need maximum liquidity and acceptance. wstETH is Aave's third-largest collateral asset (as of late 2025, accounting for two-thirds of all LST lending deposits on Aave). The stETH/ETH Curve pool routinely clears nine-figure swaps with under 10bps slippage.

How it actually works

Deposit ETH → Lido contracts route it to vetted node operators → you receive stETH at 1:1, which rebases daily to reflect staking rewards. Wrap to wstETH for a non-rebasing version — same exposure, but price-up rather than balance-up, which most DeFi protocols and institutional accounting workflows prefer.

Lido V3 / stVaults (January 2026)

On January 30, 2026, Lido V3 launched on Ethereum mainnet, introducing stVaults — isolated, modular staking environments where institutions, L2s, and node operators can run custom validator configurations, set their own risk parameters and fees, and optionally mint stETH while retaining Lido's DeFi liquidity. Linea integrated stVaults on day one for native yield on bridged ETH; Nansen launched an institutional staking product via stVaults the same week. The Lido infrastructure fee is 1% (waived to 0% for vaults over 250 ETH through March 31, 2026). This is a structural shift from Lido V2's single pooled model to a composable staking infrastructure platform.

Trade-offs

Most centralised of the major plain LSTs. The curated operator set numbers ~50 vetted entities, supplemented by the Simple DVT module (Obol + SSV Network) adding distributed validator capacity. Lido's market share has declined from 32% (2023) to ~24% (May 2026) without direct intervention — partly through competition, partly through protocol-level changes. The recurring "Lido above 33%" governance-attack concern remains, and stETH carries regulatory tail risk no other LST does at this size.

Yield

~2.6% APR (net of 10% protocol fee; Lido splits fee equally between node operators and the DAO Treasury). The 10% fee is the lowest among centralised or semi-centralised staking providers.

The bear case

If Ethereum's validator economics shift materially (MEV burn proposals, validator-size caps, PBS changes), Lido's large legacy contract surface area is the hardest to migrate. Revenue fell ~23% year-on-year as staking competition compressed fees and yields.

Rocket Pool (rETH) — deep dive

Full guide: Rocket Pool.

Rocket Pool is smaller and slower than Lido, and the most decentralised major LST — its Saturn One upgrade in February 2026 meaningfully reduced operator capital requirements.

Best for

Decentralisation-conscious stakers, and anyone who wants a non-rebasing LST out of the box. Saturn One (launched February 18, 2026) introduced MEGAPOOL validators requiring just 4 ETH per validator (down from 8 ETH under the LEB8 system and 16 ETH originally). RPL collateral is no longer mandatory — RPL is now optional and used to boost node-operator commission rates rather than as a required bond. Over 3,200 independent node operators run the network as of Q1 2026, following a 35% surge in active nodes post-Saturn.

How it actually works

Deposit ETH → receive rETH at the prevailing exchange rate (rETH/ETH greater than 1, growing). Solo node operators now contribute 4 ETH per MEGAPOOL validator alongside optional RPL collateral; the protocol sources the remaining 28 ETH from pooled depositors to fill a 32-ETH validator slot. The mismatch between operator stake and depositor ETH creates the yield premium.

Trade-offs

TVL contracted sharply through 2025–2026 — from a ~$3.5B peak to approximately $854M–$1.1B as of May 2026, as yields compressed and institutional flows favoured stETH and Ether.fi. DeFi liquidity is thinner: large rETH↔ETH swaps move price more than equivalent stETH swaps. The Saturn tokenomics rework also introduced RPL governance uncertainty.

Decentralisation

~3,200 permissionless node operators. No single entity controls above 5% of Rocket Pool's validator stake. rETH held within 1% of its underlying exchange rate even through June 2022's forced-seller crisis.

Yield

~2.1–2.3% APR per vaults.fyi (May 2026). Slightly below stETH on current data — the prior ~3.4% figure predates yield compression and Saturn's revised commission economics.

Ether.fi (eETH / weETH) — deep dive

Full guide: ether.fi.

Ether.fi is the largest LRT by TVL and now rivals Lido in total size. It is fundamentally a different risk profile from a plain LST — treat it as a restaking position, not a staking one.

Best for

ETH holders who want base staking yield plus EigenLayer restaking rewards in one non-custodial position. Ether.fi is non-custodial — depositors retain validator keys via NodeOperators-as-a-Service, which distinguishes it from Lido and Rocket Pool.

Trade-offs

Newer protocol (launched 2023). Adds restaking complexity on top of base staking — you inherit EigenLayer's slashing rules plus those of every AVS the protocol opts into. weETH bridging is being deprecated on several low-TVL chains effective June 30, 2026. If you don't know what an AVS is, you probably want plain stETH instead.

TVL and growth

$7–9B TVL as of May 2026 — now the second-largest LST/LRT by size, exceeding Rocket Pool, Frax, and Coinbase combined. Fastest-growing since 2024.

Yield

~3.0% base staking + restaking rewards (variable, depends on AVS demand) + ETHFI token incentives. Realised cash yield is roughly 3.0–4.0%; the token incentive component is speculative.

stETH vs rETH vs weETH: the comparison most guides skip

Three different risk profiles. stETH is the DeFi reserve currency. rETH is the decentralisation play. weETH is an LRT pretending to be an LST — treat it like a restaking position, not a staking one.

Best LST by use case

stETH for DeFi composability, rETH for decentralisation, sfrxETH for programmatic yield, weETH for restaking, cbETH for institutional, Pendle PT for fixed yield. Last verified: 2026-05-27.

• Best LST for DeFi composability — stETH / wstETH (Lido). Universally accepted as collateral; wstETH is Aave's third-largest collateral asset.
• Best LST for max decentralisation — rETH (Rocket Pool). 3,200+ permissionless operators post-Saturn One.
• Best LST for programmatic yield — sfrxETH (Frax). ~2.8% net yield; frxETH V2 adds NFT redemption tickets for orderly exits.
• Best LST for institutional treasury — cbETH (Coinbase, NYDFS-regulated) or wstETH via stVaults (Lido V3, January 2026).
• Best LST for restaking exposure — weETH (Ether.fi). Largest restaking TVL, non-custodial keys.
• Best LST for fixed yield — Pendle (Pendle is a DeFi protocol that splits yield-bearing tokens into principal and yield tokens, allowing users to lock in fixed yields) PT-stETH or PT-rETH (lock yield 3–12 months).
• Best LST for Curve LP — stETH/ETH pool (deepest on the protocol).
• Best LST to lever on Aave — wstETH (non-rebasing version preferred for accounting; Aave's 3rd-largest collateral asset).
• Best LST for first-time stakers — cbETH if you already use Coinbase, stETH otherwise.
• Best LST to avoid — Any LST under 12 months old with under $100M TVL or no recent audit.

How we evaluate an LST (the four-step rubric)

Operator decentralisation, depeg history under real stress, DeFi liquidity depth, withdrawal mechanism. An LST has to pass all four before it's worth using as core ETH exposure. Last verified: 2026-05-27.

1. Operator decentralisation. How many independent validator operators secure the LST, and how easy is it to become one? Lido's curated set (~50 entities) is professional but small, supplemented by the Simple DVT module. Rocket Pool's 3,200+ permissionless operators via Saturn One's 4-ETH MEGAPOOL model is the current gold standard. Anything with under 10 operators is one slashing event from a bad day.
2. Depeg history under real stress. Has the LST been tested through a forced-seller event? stETH hit $0.93 in June 2022 (Celsius/3AC unwinds) and recovered cleanly. rETH stayed within 1% throughout. Post-Shapella (April 2023), direct ETH redemption keeps LST-to-ETH arbitrage tight — stETH has traded within 0.2% of ETH since mid-2023. Newer LSTs that haven't been through a stress event are pricing zero crisis risk — that's free option value going the wrong way.
3. DeFi liquidity depth. Can you swap $10M of the LST back to ETH without moving the price more than 50bps? stETH yes, rETH yes-ish (TVL contraction in 2025–2026 thinned its secondary liquidity), most others no. If you might need to exit in size during stress, only stETH reliably supports it.
4. Withdrawal mechanism. Does the LST support direct ETH redemption from the protocol (all major LSTs do post-Shapella), and what's the queue length? Lido routinely processes withdrawals within 1–5 days; smaller LSTs can take weeks depending on validator exit queues. Always test a small withdrawal before sizing a position.

Risk summary

Four risks: smart-contract (audited but nonzero), depeg (stETH hit $0.93 in 2022, stable since Shapella), slashing (under 0.05% historical), regulatory (SEC softened post-2024 on decentralised LSTs). Last verified: 2026-05-27.

• Smart-contract risk. All major LSTs have multiple audits but nonzero risk. Lido V3's stVaults architecture adds new contract surface area — modular vaults, operator tier logic, and minting caps — on top of the already-complex withdrawal and dual-governance system. Rocket Pool's Saturn One MEGAPOOL contracts were audited pre-launch but introduced new economic logic (optional RPL, per-validator revenue splits) that is less battle-tested than the V1 minipool model.
• Depeg risk. stETH dropped to $0.93 during the June 2022 Celsius/3AC crisis — not because Lido failed, but because forced sellers blew through Curve liquidity. Since Shapella withdrawals (April 2023) enabled direct redemption, arbitrageurs keep stETH within 0.2% of ETH under normal conditions. ezETH (a restaking LRT) briefly hit $0.74 in April 2024 when Renzo's airdrop allocation announcement triggered exits faster than redemption queues cleared. No major new plain-LST depegs occurred in 2025 or 2026. Assume any LST can wobble during forced liquidation cascades.
• Slashing risk. Diluted across operators. Historical slashing on major LSTs is under 0.05% of total staked. The bigger systemic risk would be a correlated-slashing bug in a dominant consensus client — a theoretical worry that has never materialised at scale.
• Regulatory risk. Kraken settled with the SEC in February 2023 ($30M, shut down US staking). The SEC's enforcement posture on ETH staking-as-a-service softened materially through 2024–2025 amid crypto-friendly regulatory shifts; Coinbase's staking program continued operating. Decentralised LSTs (Lido, Rocket Pool) and self-staking carry the least regulatory beta; centralised-exchange staking products remain the highest exposure.

Realistic LST portfolio (May 2026)

Sample $25k ETH allocation: $15k wstETH on Aave, $5k Pendle PT-stETH, $5k rETH — keeps ETH exposure, layers fixed yield, and diversifies validator-operator concentration. Last verified: 2026-05-27.

For a $25,000 ETH allocation:

Keeps ETH price exposure + base staking + a fixed-yield slice + operator-concentration diversification. Blended net yield roughly 2.5–3.0%. The wstETH-on-Aave position also unlocks 80–85% LTV borrowing capacity if you want to lever the position (Aave E-Mode for ETH-correlated assets).

Looking ahead to 2027

Specific things worth tracking over the next 12 months:

• Lido market share. Share fell from 32% (2023) to ~24% (May 2026) without a hard cap — driven by restaking competition, institutional ETF staking, and centralised-exchange growth. If Lido's share continues falling, stVaults become the growth lever; if it stabilises above 30%, the Ethereum social-consensus pressure on the "33% threshold" concern escalates again.
• DVT (Distributed Validator Technology) rollout. Lido's Simple DVT module and Obol deployments are adding genuinely-decentralised validator capacity. If DVT crosses 10% of staked ETH, the LST centralisation debate fundamentally shifts.
• Saturn One impact on Rocket Pool TVL. The 4-ETH MEGAPOOL model should lower barriers enough to attract new operators. Whether that reverses the TVL outflow (~$3.5B peak to ~$854M trough) or merely stabilises it is the key 2026–2027 variable for rETH.
• Restaking-without-restaking products. Pendle PT-weETH lets you capture the eETH yield curve without the restaking-slashing exposure. Expect more "wrapper" products that strip LRT risk and resell the cash component as fixed yield.
• The next forced-seller event. stETH's 2022 depeg only happened because levered funds got margin-called. The next time leveraged ETH carry-trades unwind in size, watch the stETH/ETH Curve pool — it is still the canary for plain-LST peg stability.

---

Related: How to Stake ETH · Best Restaking Protocols 2026 · DeFi Yield Farming Guide