stETH Yield Update — Week of October 5–11, 2026

1. The week for yield, in one line

Yield is supposed to be the boring half of a staking site, and this week it was the only part of the story that went the right way. On the Ebunker series, the network base APR refreshed for the first time in two editions and printed 2.51% for the week ending October 11 — a +33bp step that restores most of September's 39bp fall and settles the question this column has carried for two editions: 2.18% was a one-week aberration, not a regime. The stETH 7-day average printed 2.53%; Lido's own October 6 statement put the stETH annualized rate at 2.23%, at or above its level since August; and the user-net band this site carries for a holder's actual return stays at 2.21–2.24% (mid-point 2.22%). Meanwhile ETH itself fell 6.9%. A yield holder is paid in ETH, so a lower ETH price buys more of it — but that is a comfort about future income, not a hedge against this week's mark.

Source: Ebunker Ethereum staking weekly (week ending Oct 11, 2026: base APR and stETH 7-day); Lido statement (Oct 6, 2026: 2.23% stETH rate); user-net band per its dated sources in the data report; ETH print per the DefiLlama coingecko feed.

2. What a stETH holder actually earns

The gap between the headline network APR and the yield a holder keeps is the whole point of a yield update, so it is worth stating on one basis. Lido's protocol fee (10% of staking rewards) takes the 2.53% stETH 7-day average down to a user-net band of 2.21–2.24% (mid 2.22%), and Lido's own October 6 print of 2.23% sits inside it. A holder of 10 ETH at the October 11 print ($2,508.04; position $25,080, DERIVED) earns about $557 a year at that mid-point. Set that against cash: the 3-month T-bill at 3.92% (Sep 1 print) returns $983; the 2-year at 4.34% (Aug 28) returns $1,088; the 10-year at 5.26% (Oct 5) returns $1,319. The carry gaps are −$426, −$532 and −$762 respectively — all three negative, and all three computed on the same 2.22% mid-point.

What changed this week is the direction, and it changed back. Last edition argued that the October 2 jobs report had, for the first time since the September hike, pointed the carry tables the right way. The September FOMC minutes released October 7 took that back: most officials, the minutes showed, still expect another rate increase before the end of 2026. The 10-year Treasury touched a 24-year intraday high near 5.34% before settling in the 5.24–5.28% area, and the 30-year pushed toward 5.68%. On the refreshed 10-year print the carry gap has widened to roughly −$762 a year. The October 28 meeting is a live question again, and until it resolves, staking's carry disadvantage is set by a central bank, not by a protocol.

Source: position value DERIVED on the DefiLlama coingecko feed Oct 11 print; staking and cash rate rows carried with their print dates; FOMC minutes (Oct 7, 2026) and Treasury levels via BlockFence and the Gate Ventures weekly recap (Oct 5, 2026).

3. The 39bp step, reversed — and why it was a print, not a policy

Two editions ago the network base APR fell 39bp in a single week, to 2.18%, the largest one-week move in this site's record, and this column refused to call it a regime without a second print. That patience is now paid off: the Ebunker weekly resumed and printed 2.51% for the week ending October 11. The series reads 2.62% → 2.57% → 2.57% → 2.18% → 2.51% across September into October — a spike down and most of the way back, which is the signature of a measurement artifact rather than a change in the protocol's payout curve. The issuance curve that sets base yield moves slowly and inversely with the square root of total stake; a 39bp single-week drop, followed by a 33bp single-week recovery, is far too fast for that curve to explain, and the honest conclusion is that one print was noisy and the other is closer to the trend.

The distinction matters for anyone who acted on the 2.18% figure. A carried register that said “no fresh read” for two editions was less exciting than a fresh number would have been, but it was the right call: had this site chased the 2.18% print with a corroborating source, it would have published a regime call that lasted one week. The resumed print also carries the caveat that a single reading re-establishes a level, not a trend — one more print near 2.5% would confirm the step back, and a print back at 2.18% would reopen the whole question.

Source: Ebunker weeklies (Sep 7, 14, 21, 28; week ending Oct 11, 2026); the 39bp step's mechanics per the earlier deep dive; issuance-curve behavior per Ethereum staking documentation.

4. BitMine's own number, and what the cap does to the yield bid

The treasury bid reports its own print, and this week it also reported its own end. BitMine's filing for the week ended October 4 listed 6,016,414 ETH, about 4.9% of the 122.1 million supply figure it uses, with roughly 84% deployed in staking and 5,053,788 ETH staked — about 11.6% of all staked ETH (DERIVED against the 43,865,190 ETH board). BitMine's own staking revenue print has run in the 2.6% area on its accounting, above the network's base APR because it reports gross of the protocol's fee split and on its own window; the two are printed side by side and never averaged. Then, on October 7, chairman Tom Lee put a 5% hard cap on holdings, roughly 88,586 ETH away.

For the yield story the cap is a two-sided fact. On one side, BitMine's staked balance — already the largest single slice this column tracks — keeps earning after the purchases stop, and Lee said staking rewards would continue adding to the position even past the cap. On the other side, BitMine has been a steady buyer of the ETH that funds new stake, and a cohort of loopers and funds has priced its bid as a standing feature of the market. Removing a recurring buyer does not remove the yield from existing stake; it removes the incremental demand that stake conversion depends on. The yield question and the price question separate cleanly here: staking rewards accrue to whatever is staked, regardless of who bought it and when.

Source: BitMine disclosure for the week ended Oct 4, 2026, and Tom Lee remarks at TOKEN2049 (Oct 7, 2026) via CoinPaprika, Coinspress and ME News; network staked total per the Ebunker weekly (week ending Oct 11, 2026).

5. The MetaMask exit: the drag estimate, updated

This is the section the last two editions pointed at, and the data has now had a week to answer it. MetaMask's precautionary exit removed roughly 523,000 ETH from earning position — about 1.19% of the 43,865,190 ETH staked network-wide (DERIVED). Exited validators earn nothing until they re-enter, so the pre-event estimate was a drag of at most about 3.0bp on the network 7-day average in a single week (2.51% × 1.19% idle, DERIVED on the refreshed base rate). The observed result: the network base APR printed up 33bp over the same window, and Lido's stETH rate held at 2.23%, which the protocol explicitly attributed to operator diversification and a shortened entry queue. In other words, the drag is real but smaller than the noise in the print that would carry it — you cannot see a 3bp headwind in a 33bp up-move, and pretending otherwise would be reading tea leaves.

The re-entry side is now measurably easier than the 45-day estimate implied. The entry queue shortened from about 1,480,361 ETH (Oct 4) to 1,379,437 ETH (Oct 10), cutting the wait to 23 days 23 hours; the exit side rose to 964,365 ETH with a 16-day 18-hour wait. With the ~7.4-day withdrawal sweep, the round trip sits near 24 days in both directions — shorter than the 45-day worst case Lido published on September 30, because that estimate was built on a longer entry queue than the one that materialized. For a yield holder the practical read is that the rewards gap from this incident is closing faster than feared, and the network print never registered it at all.

Source: exited stake per Lido's Sep 30 disclosure and Oct 6 completion statement; network staked total and base APR per the Ebunker weekly (week ending Oct 11, 2026); queue levels and waits per validatorqueue.com via CryptoCompass and CryptoTicker (Oct 10, 2026); drag and round-trip DERIVED.

6. Issuance and burn: the small negative under the yield

Staking yield is not the whole return — ETH's own supply change sits underneath it, and that line refreshed this week too. The Ebunker weekly's issuance window showed net issuance of +20,214.32 ETH over seven days, about +0.017% of that series' supply denominator, annualizing to roughly 0.87% (DERIVED) — essentially unchanged from the carried 0.86%. Against that, the burn side keeps the carried pair that frames it: a 30-day average base fee of 0.179 gwei against the 13.384 gwei needed to hold supply flat, a ratio of roughly 1/75. The implication for a yield holder is the same every week: staking pays ~2.2% net while ETH supply grows ~0.87%/year, so the real carry on staked ETH is the staking yield minus the dilution, and the cash leg still wins on the nominal comparison this site runs. The issuance figure is now a fresh in-window print rather than a carried one; the burn figures remain thresholds, not this week's prints, and are labelled that way.

Source: Ebunker weekly (week ending Oct 11, 2026) for the issuance window; carried base-fee and breakeven figures per their Sep 2 print dates; annualization DERIVED.

7. Method and what was withheld

The supply cross-check stays withheld. The stETH supply input behind the supply×price comparison is dated September 9 — over a month old — so rather than run a stale input as current, the data report keeps that check suspended and derives the same quantities (ETH under Lido management, share of stake) from two in-window prints instead. This update uses those derived figures and labels them as derived. Every yield number here traces to the Ebunker weekly that closes this window, to Lido's October 6 statement, or to in-window price prints.

Source: site_data.py (Snapshot #8, as of Oct 11, 2026) and the edition archive; supply-withhold note per the data-report ledger.

8. What to watch for yield

Four markers, all checkable on dated prints. (1) The Ebunker level — whether the resumed 2.51% print holds or reverts; one more print near 2.5% confirms the step back from 2.18%, a return to 2.18% reopens it. (2) The MetaMask drag — a measurable dip in the network or stETH 7-day print attributable to the exit would put a basis-point price on the incident for the first time; so far the print has moved the other way, which is itself the finding. (3) The BitMine pivot — whether a buyer that will soon stop buying leans harder on staking rewards; a move in its staked balance off ~5.05M would be the signal, and its 11.6% share makes it a real input to the network average. (4) The October 28 FOMC — the meeting this week's minutes reopened; a hold narrows every cash leg, a hike widens the gap once more, and no staking yield has to change either way. The yield story next week is whether the refreshed board holds and whether the macro leg stops moving against the carry trade.

Source: markers constructed from the figures in sections 1–6; the watchlist is the yield half of the intelligence scorecard.

FAQ

Related reading

• Lido & stETH Weekly Data Report — Week of October 5–11, 2026
• The Steadiest Buyer: What BitMine's 5% Hard Cap Ends
• Weekly Intelligence — Week of October 5–11, 2026