LSTs vs Restaking in 2026 — Which Actually Yields More? stETH, rstETH, eETH, rsETH Compared
Open any staking dashboard today and you’ll see numbers like stETH 3.1% APY, rstETH 5.4%, eETH 6.8%, rsETH 7.2%. Many people pile into rsETH on reflex — but these four numbers are not on the same scale. The first two already net of protocol fees; the last two are before AVS slashing exposure. The first two carry only Ethereum’s own risk; the last two add EigenLayer/AVS smart-contract risk. This post lays the four tokens side-by-side along three axes — yield layers, risk layers, and liquidity discount — so the next APY you read gets discounted correctly in your head. If “restaking” is still fuzzy, start with restaking and EigenLayer explained and what is liquid staking.
What each token actually is
Mapping names to stacks first.
| Token | Protocol stack | One-line |
|---|---|---|
| stETH | Lido | The largest pure LST, plain ETH staking |
| rstETH | Lido + Lido’s Modular Restaking module | stETH plus Lido’s CSM/node extension for a thin extra layer |
| eETH | Ether.fi | LRT (Liquid Restaking Token) — ETH staked and pushed into EigenLayer |
| rsETH | Kelp DAO | LRT — ETH routed through EigenLayer into multiple AVSs |
Loosely: stETH is one layer, rstETH 1.5, eETH and rsETH two layers. Each extra layer stacks new contract risk, new incentive sources, and new slashing rules.
Layer 1: base staking yield
LST or LRT, the underlying ETH ultimately sits on the Ethereum beacon chain — that’s the PoS base yield. In 2026 issuance + priority fees + MEV land around 2.9% - 3.2% APY. That is the shared denominator of every ETH yield token; everything else is markup on top.
- stETH stops here and you pocket ~3.1%
- rstETH, eETH, rsETH “reinvest” this layer, so their headline APY = base + upper-layer incentives
One often-overlooked deduction: Lido charges 10% on node operator + protocol fees, so stETH nets ~90% of the underlying. Ether.fi and Kelp have slightly different fee schedules but the magnitude is similar.
Layer 2: where the extra yield comes from
This is where LSTs and LRTs diverge.
- rstETH upper layer: Lido’s own Community Staking Module plus liquidity incentives — essentially Lido subsidizing early users with a mix of LDO emissions plus marginal native ETH. Adds roughly 1.5% - 2.5%.
- eETH upper layer: Ether.fi parks the deposit across EigenLayer and Symbiotic operators while emitting ETHFI tokens plus future EIGEN airdrop expectations. Adds 3% - 4.5%, but a large chunk is token “expected value”, not steady cash flow.
- rsETH upper layer: Kelp delegates through EigenLayer into multiple AVSs, each paying restakers. Adds 3% - 4% with KEP/EIGEN tokens being most of it.
So a 7% APY is not “7% in ETH” — it’s typically “3% ETH + 4% protocol token (priced in ETH today)”. If those protocol tokens fall, your ETH-denominated return falls right with them. This is the prediction gap rookies miss.
Layer 3: slashing exposure is not equal
Slashing risk is where the real cost of LST/LRT diverges.
- stETH slashing exposure comes only from Lido operator misbehavior on mainnet. Historical losses < 0.002%, essentially negligible.
- rstETH adds CSM (community-node) slashing; theoretically a tick higher but historically still very low.
- eETH / rsETH add an AVS slashing layer — every AVS has its own slashing rules, and those rules are still mutating throughout 2026. One AVS going bad can in theory slash 20%-100% of the ETH delegated to it via EigenLayer.
That’s the real hidden cost. A 7% headline looks great until you realize a 5% “AVS black-swan probability” with a 30% principal loss makes your expected return negative. For AVS state of play, see EigenLayer AVS status 2026 and real restaking risk cases — I documented every incident I could find.
Liquidity discount: can you actually swap 1:1 back to ETH?
A yield token is not ETH. Secondary markets often trade at a discount, and that discount is a realized loss the moment you exit.
| Token | 2026 average discount | Extreme |
|---|---|---|
| stETH | 0.05% - 0.2% | Down to 7% in May 2022 |
| rstETH | 0.1% - 0.5% | New token, limited extreme samples |
| eETH | 0.2% - 1.0% | Liquidity stress can spike 2-3% |
| rsETH | 0.3% - 1.5% | LRT cohort had a 5% collective discount in May 2024 |
Whatever discount you sell at gets deducted straight from APY. A 2% discount on a 7% APY token erases roughly 30% of one year’s return. So any “I’ll never sell” thesis must include a real lock-up tolerance — discounts shrink to footnotes if you genuinely won’t exit for a year. Buying LRTs as a “high-yield checking account” makes the discount your primary risk.
Putting all axes on one table
| Token | Headline APY | Net of token volatility | Slashing expected | 6-mo liquidity discount | “Realistic lock-in APY” |
|---|---|---|---|---|---|
| stETH | 3.1% | 3.0% | -0.01% | -0.1% | ~2.9% |
| rstETH | 5.4% | 4.0% | -0.05% | -0.3% | ~3.6% |
| eETH | 6.8% | 4.5% | -0.4% | -0.6% | ~3.5% |
| rsETH | 7.2% | 4.6% | -0.5% | -0.9% | ~3.2% |
Estimates plus personal modeling — not investment advice. But notice the counter-intuitive takeaway: after stripping token volatility, slashing expectation, and liquidity discount, the four products’ realistic lock-in APYs cluster around 3%. That makes sense — markets don’t leave a 3% free lunch between four heavily traded products. “Higher yield” is paid for in more complex risk surface.
So which one to hold?
My personal framing by capital tier:
- Core, long-term: stETH. Cleanest structure, deepest liquidity, accepted as collateral on Aave and the rest of DeFi.
- Satellite, comfortable with extra protocol risk: rstETH. ~0.5% extra realistic APY over stETH at the cost of one Lido-internal layer.
- Higher beta on EigenLayer thesis: eETH. Ether.fi’s execution and ETHFI secondary liquidity lead the LRT field.
- Airdrop and AVS exposure play: rsETH. I cap this at 5% of position size, the tail of multi-AVS slashing remains under-priced.
The four are not substitutes, they’re complements. Sizing across them beats betting everything on the highest headline APY.
Revisit in a year
I’ll re-run this comparison in May 2027. The real verdict on LRTs comes not from 2026’s APY but from looking back in 2027 at how many AVS incidents occurred, how many tokens unlocked, how much incentive tokens devalued. Yield comparison is a moving game, not a snapshot. Until then, stETH stays my anchor and the other three are “satellites” whose combined size sits under half of my stETH allocation.