stETH Yield Update — Week of August 17–23, 2026

1. The corridor this week

The founding yield snapshot for the week of August 17–23:

The spread between the ~2.6% gross and the 2.17% realized is fully accounted for: ten points of protocol fee, plus smoothing and variance across a 600+-operator set. Nothing is leaking; the corridor is simply lower than it was — 2.51% in January on the same DefiLlama series.

Context for why the corridor, not the point reading, is the right object to track. A single week's APY is noisy — MEV dispersion alone can swing realized rates by tens of basis points between operators. The corridor (currently ~2.17% realized to ~2.6% gross on verified reads, with 2.0% as the scenario floor from our deep dive, not an observed print) moves on months-long drivers: validator count, fee regimes, participation. Tracking the corridor's edges catches regime changes early; tracking weekly point prints mostly catches noise. That is why this column's headline metric will be the corridor width and midpoint, not Monday's APY.

The band arithmetic, restated on the corrected price basis so it survives the correction banner: at ~$2,500/ETH, the corridor's plausible span (2.0% scenario floor to 2.4% bull end, both labels from our deep dive rather than observed prints) earns roughly $500–$600 per 10 ETH annually (DERIVED) — a ~$100 spread at the round basis, versus the $74 the founding draft printed at the stale quote. The founding week's lesson generalizes: a mispriced ETH quote does not just misstate one dollar figure, it misstates every band, gap and example built on it. Worked examples now carry their price basis in the same sentence as the rate — a discipline this correction paid for.

2. Why the net number matters more than the gross

Marketing pages quote gross; portfolios earn net. After Lido's fee, a perfect validator earns ≈2.34%; the protocol-wide realized print of 2.17% reflects real-world penalties and reward timing. When comparing any two staking products, insist on same-source, same-date net figures — the 2.17% vs 2.18% stETH/rETH parity (Aug 15) is the cleanest example: at parity, neither protocol can win your capital on rate alone.

3. Competitor rate board

• Rocket Pool (rETH): 2.18% — parity with stETH; compression is systemic.
• Custodial exchange products: Binance Staked Ethereum sits at $7.00B TVL (DefiLlama, Aug 14); rate quotes for custodial products vary by platform and tier, which is precisely why we do not average them into one board number.
• Restaking overlays: additional AVS rewards are often marketed atop the compressed base, but no public blended-rate figure verifies what is actually being paid this week (sector research, Aug 2026). Treat as optionality, not yield.

4. Peg & redemption watch

No redemption-queue stress surfaced in whitelisted sources this week, and the derivatives complex was orderly (global ETH futures OI ~$112.6B, +1.09%; benign funding; balanced long/short — Coinglass, Aug 27). Historical pattern: peg pressure arrives with leverage spikes and crowded funding, not with orderly books. The stETH/ETH secondary market printed no notable discount in any source we checked; formal discount tracking starts next issue with our second snapshot on file. Queue depth itself gets a standing table from the same date — we will not quote queue numbers we have not verified.

One more calibration while the queue table is still pending: this section's null result is only as good as its coverage. "No stress surfaced in whitelisted sources" means the sources we check showed nothing — it does not mean the whole market was measured. The derivatives read (OI ~$112.6B, benign funding, Aug 27) is the best proxy available because leverage is where peg stress historically incubates, but the honest-missing list — queue depth above all — is what turns this section from anecdote into measurement. It is scheduled, not yet delivered, and this paragraph exists so the gap is not forgotten once the tables arrive.

The peg mechanics deserve one more sentence than they usually get. stETH's price is held near parity by a redemption curve that slightly discounts exits: a staker leaving pays a spread that widens with queue depth, which is why the token trades a hair under ETH in calm markets and why the number to watch is not the price itself but the spread's behavior under stress. A peg that holds at 0.999 in a flat week tells you little; a peg that holds through a redemption spike tells you the mechanism works. This week saw no such spike, so the honest reading is "no new information" — a null result this column will print as such rather than dress up as reassurance.

5. LDO market context

LDO spent the window inside a $0.29–0.37 band (daily logs, Aug 13–22), recently ~$0.336 with a ~$301M market cap (Aug 27) — a token trading on macro risk appetite and expectations for Lido's future role rather than on cash flows. With yields at parity across protocols, the investment case concentrates on share of stake, product expansion (the new Earn vault layer) and the pending US ETF form factor.

Band position for the token, since the headline number is a point: $0.336 sits just above the midpoint of the $0.29–0.37 band — 57% of the way from floor to ceiling (DERIVED), and 9% below the ceiling. With yields at parity across protocols (the board above), there is no rate-based case for one LST over another this week, so the token trades on exactly the three factors listed above — share, product, form factor — and on the macro calendar's next print. When the rate differentiator is zero, the non-rate factors carry all of the differentiation; that is the cleanest statement of LDO's setup at the founding snapshot.

6. The week ahead

Three markers for next week's update: the post-PCE risk backdrop, any VanEck filing movement, and — our own milestone — the first true week-over-week yield delta once the second snapshot lands. The monthly consolidated report (Liquid Staking Market Report — August 2026) follows at month-end.

7. A worked example: 10 ETH through the corridor

Derived arithmetic, clearly labeled as such. Ten ETH staked through Lido at current readings:

• At the gross network yield (~2.6%): 0.26 ETH over a year — before anyone's fees.
• After the 10% protocol fee: ≈0.234 ETH — the perfect-validator net.
• At the realized stETH APY (2.17%): ≈0.217 ETH — about $400 a year at the ~$1,845 snapshot price.

The ~0.043 ETH total gap between gross theory and realized practice decomposes cleanly: the 10% protocol fee accounts for ~0.026 ETH, and smoothing plus variance across the 600+ operator set — missed attestations, MEV dispersion, occasional penalties — for the remaining ~0.017 ETH. It is not hidden fees; the split is on-chain and published. Why this matters for planning: at the corridor's bear end (2.0%), the same 10 ETH pays ~0.20 ETH; at the bull end (2.4%), ~0.24 ETH. The entire plausible annual band spans about $74 per 10 ETH — which is exactly why sophisticated holders choose LSTs on liquidity and integration, not on hundredths of a percent.

Two conventions hide in the corridor table and they are not interchangeable. APY compounds — a rate earned on rewards already credited — while APR does not, so the same underlying payment prints lower as APR. At these magnitudes the gap is small but not zero: roughly four basis points separate a 2.17% APY from its APR equivalent at annual scale. Fee conventions diverge even more: a 10% protocol fee on Lido's gross, a 25% fee on cbETH's, a 14% fee on rETH's, each taken off a different gross base. Any league table that quotes one provider net and another gross will rank them wrongly. The order of operations in this column is fixed for that reason: gross first, fee second, compounding last — and the label says which convention each number uses.

8. Corridor history, on one line each

From next week, this table becomes a rolling eight-week series with WoW deltas — the founding anchors above are the fixed reference points every future print will be measured against.

FAQ

Related reading

• Weekly Intelligence — Week of August 17–23, 2026
• The Quiet Squeeze: Why stETH Yields Fell From 2.51% to 2.17%