Lido & stETH Weekly Data Report — Week of August 17–23, 2026

1. Executive summary

The founding edition of this column establishes the baseline dataset everything else will be measured against. Three findings define the current state of Lido and stETH:

• Ethereum staking is at record saturation: 41.89M ETH staked = a 34.7% staking ratio (all-time high, Aug 13), with ~789K active validators and network inflation down to 0.86%.
• Yields keep compressing: stETH paid 2.51% APY on January 5 and 2.17% by August 15 (DefiLlama); rETH reads essentially identical (2.18%), confirming system-wide dilution rather than a Lido-specific problem.
• Lido's position is intact and internally consistent: 9.63M ETH through the protocol implies ≈$17.7B in stETH value, matching DefiLlama's $17.9B TVL within rounding — a two-source cross-check that passes.

2. Baseline table — 21 metrics, each with its source date

Sources per row are listed below and linked in the Sources section. Snapshot methodology: each metric keeps its own verified date; nothing is interpolated.

3. Yield mechanics: reconciling 2.6%, 2.3% and 2.17%

Three different-looking yield numbers circulate, and they are not contradictory — they measure different things. The network-level yield (~2.6%, Aug 13) is the gross return on staked ETH before anyone's fees. Subtract Lido's 10% fee share (protocol docs) and you get ≈2.34% net for a perfect validator. DefiLlama's 2.17% APY for stETH (Aug 15) reflects realized rewards including occasional penalties, timing effects and MEV variance across the whole operator set. The lido.fi frontend rounds its rolling estimate to 2.3% APR. All three sit comfortably inside one consistent picture; quoting one without the others is how misleading staking comparisons get made.

4. Market & positioning context

Global ETH futures open interest was approximately $112.6B (+1.09% on the day) with benign funding and roughly balanced long/short ratios (Coinglass, Aug 27). For stETH specifically, the meaningful secondary-market datum remains the peg: with leverage orderly, the historical pattern is discount-free chop rather than stress. LDO traded roughly $0.29–0.37 over Aug 13–22 on daily logs, coiling ahead of the Aug 26 PCE/GDP macro prints (MetricsHour).

5. Anomalies & data-quality notes

• Aggregator divergence observed during research: one major retail platform quoted ETH at $1,571 while consensus quote engines showed ~$1,84x the same week. Conclusion for our methodology: every price on this site gets double-sourced before publication; no single-aggregator quotes.
• Validator counts differ by definition: ~789K refers to the active validator set (Staking Rewards); other trackers publish higher totals including offline/queued entities. We label ours clearly rather than averaging apples with oranges.
• Market-share arithmetic carries a date warning: 9.63M ETH through Lido (late Aug) divided by 41.89M total staked (Aug 13) produces ≈23% — close enough for context, but both inputs must be re-read on the same day before claiming a precise share or trend. Flagged as DERIVED everywhere on this site.

6. Four-week direction — the honest version

A four-week trend table normally lives here. On founding issue, I won't manufacture one. What verified anchor points do support:

• Yield compression is real and documented: 2.51% (Jan 5) → 2.17% (Aug 15), DefiLlama's continuous series.
• Lido TVL stability at the top: $17.86B (Aug 14) → ~$17.9B (late-Aug reads) — flat-to-slightly-up while total DeFi slipped.
• Staked ETH hit its ratio ATH during August, and inflation drifted lower with it (Aug 13 readings).

Next issue this section becomes a true week-over-week table once second snapshots land.

7. Conclusion: what did we learn?

Three things worth keeping. One: Ethereum staking entered August at record saturation — 34.7% locked — which mechanically guarantees further yield dilution and makes Lido's share of an ever-larger pie the number to watch. Two: every credible yield source, from DefiLlama's series to competitor parity at ~2.17–2.18%, tells the same compression story, so future issues can track one signal rather than many. Three: the data pipeline works — the Lido TVL/supply/price triangle reconciles within rounding across independent sources, which is the baseline trust property a data column needs before it deserves readers.

8. The peer board, for context

Where Lido sits among the protocols that matter, per the latest full DefiLlama weekly (Aug 14):

Two readings worth internalizing. First, the top five is really two staking businesses, two lending markets and one infrastructure layer — Lido competes directly only with the fifth row and the unlisted long tail, which is why our competition sections track custodial products as seriously as DeFi-native rivals. Second, Aave's position matters to stETH holders indirectly: lending markets are where stETH goes to work, so Aave's utilization and rate dynamics feed back into effective stETH yields. We will track that linkage explicitly once weekly deltas start.

9. A worked example: what 2.17% actually pays

Numbers this small are hard to feel, so here is the arithmetic, fully derived from sourced rates. Take 10 ETH staked through Lido at the August 15 reading of 2.17% APY: over a year at a constant rate, that position accrues roughly 0.217 ETH in rewards — about $400 at the ~$1,845 snapshot price. The same 10 ETH at the ~2.6% gross network yield would theoretically pay 0.26 ETH before fees; the 10% protocol fee and realized-variance effects bridge the difference. Two honest caveats: the APY floats week to week (it fell 34 basis points since January), and rewards compound daily through the rebase rather than arriving as a year-end lump. We publish this example so the stakes of the corridor debate in section 3 are concrete: the difference between the 2.0% and 2.4% ends of the plausible band is ~$74 a year per 10 ETH — small per unit, decisive in aggregate.

A founding baseline earns its keep only if later editions are willing to be measured against it — and, where it was wrong, corrected against it. Every week-over-week figure this column will ever print is computed back to the numbers on this page. That is why each row carries its source and snapshot date rather than a bare value: a delta without a dated baseline is not a measurement. Where later verification finds a baseline figure was wrong, the correction is stamped onto this page and the corrected value is used going forward — the archive keeps the original, the ledger keeps the reason. Treat this page the way a statistician treats a first print: load-bearing, dated, and revisable.

10. What changes in next week's edition

• True WoW deltas for every baseline metric, once the second same-day snapshot is on file — the table in section 2 gains a change column and the four-week window opens.
• Same-day market share: Lido ETH and total staked ETH read from matched dates, replacing the current ≈23% derived placeholder with an auditable figure.
• Standing quality flags: any metric whose two sources fail to reconcile gets a visible flag in the table instead of a footnote.

That last one is the column's real promise: the dataset improves in public, and nothing gets quietly restated.

FAQ

Related reading

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