stETH Yield Update — Week of August 31–September 6, 2026

1. The corridor, on its widest source base yet

For the first time every layer of the stack sits on one page with its source named — and six of the seven rows carry a date. The exception is the Lido frontend, which publishes a rate with no date and no stated basis; we flag that below rather than paper over it. The honest headline: a stETH holder is earning somewhere between 2.21% and 2.40%, and the spread between those two numbers is a measurement problem, not a choice anyone can make.

Why the range rather than a number: a 2.60% protocol rate net of Lido's 10% fee should read ~2.34%, which is above the 2.21–2.24% band actually observed — implying the observed band covers a slightly weaker rewards window. The frontend's 2.40% states no basis at all. We are not going to average our way to a clean-looking figure. What the range does tell you unambiguously: every credible reading of a stETH holder's net yield now sits 152 to 171 basis points below short-dated risk-free cash.

2. The carry math, against two tenors

On a 10 ETH position — $24,799 at the Sep 6 close of $2,479.92:

• Staked via Lido at the observed net mid (2.22%, the rounded midpoint of the 2.21–2.24% band): ~$551 a year.
• At the frontend's 2.40%: ~$595 a year. The $44 difference is the basis ambiguity, not an option anyone has.
• The same capital in 3-month bills at 3.92%: ~$972 a year — a gap of ~$422.
• Or in 2-year notes at 4.34%: ~$1,076 a year — a gap of ~$526.

So the annual cost of holding staked ETH rather than cash is roughly $422 to $526 per 10 ETH, and the two ends of that range are a tenor choice, not an uncertainty. What you are buying with it is ETH upside and on-chain composability. That is a legitimate purchase — but it is a purchase, and this week's numbers price it more explicitly than at any point in this site's history. Note also that the gap is narrower than last week's $545 — but not for the reason that comparison invites. Last week's figure used a ~2.16% derived net rate on a ~$25,000 position ($540 against $1,085); this week's uses the 2.22% observed net rate on $24,799 ($551 against $1,076). Of the ~$19 narrowing, roughly $15 is that change in which net rate we can source and only about $4 is ETH's price (DERIVED: $25,000 × 2.18% = $545, against $24,799 × 2.18% = $541; the remaining ~$15 is the move from a 2.16% basis to a 2.22% basis on the same position). Neither is a yield improvement: the two net rates come from different sources on different bases, and a verified change in what stakers earn would need two prints from the same source on the same basis. We do not have that yet.

3. Parity restored — the call from edition one, finally testable

This column has carried an unresolved parity question since the founding edition: do stETH and rETH yields actually track each other, or was that an artifact of a single source? This week both are readable on a comparable basis from the same third party. stETH net 2.21%–2.24%; rETH 2.18%–2.20%. The gap is 1 to 6 basis points. Parity holds.

One caveat we owe you: last week's frozen cross-provider table put rETH's net at 1.72%, which is nowhere near 2.19%. That table and this week's source disagree by roughly 47bp on the same asset. We are not averaging them. The new source is used because both protocols appear in it on a stated basis; the old table is retained in the archive with its date.

4. LDO's +11% week

LDO went from $0.3527 (Aug 30) to $0.3917, +11.1%, with the move concentrated in a +8.2% session on Sep 3. Coverage attributed it to a September protocol update reassessing Lido's risk-and-reward profile; the same coverage explicitly noted no new products, regulatory developments or ecosystem events accompanied it. Weekly range $0.3447–$0.4050, with resistance identified at $0.4203. Two other trackers printed ~$0.40 and $0.4131 the same day — a ~5% spread we flagged in the data report rather than resolving by picking one. Last week LDO fell 12% on a hawkish Fed; this week it rose 11% on a PDF. Both weeks, the staking business did nothing.

5. Peg & redemption watch

stETH $2,480.62 against ETH $2,479.92 — a ratio of 1.0003, no discount, and the second consecutive edition with a measured peg read on file: the founding edition's check was qualitative, and edition two quoted ETH as a range rather than a same-day close. The structural reason is now measurable for the first time: the exit queue is empty and exit wait is under three hours. Peg stability is a function of exit capacity more than of confidence, and exit capacity is currently abundant. Redemption-queue depth for stETH specifically remains on our honest-missing list — the network-level exit queue is a proxy, not the same metric, and we will not present it as if it were.

What would actually break the peg, stated plainly: a discount needs a seller with no working exit. Right now neither half of that condition is close — the network exit queue is empty, exit wait is under three hours, and stETH's secondary market is the deepest in the sector. The setup worth watching is a leveraged unwind that forces on-chain redemptions faster than the queue can clear. That scenario is laid out in this week's deep dive, and it is not currently visible in any data we can verify. We would rather say that than imply the peg is safe because it looked safe twice.

6. Corridor history — the anchors

The through-line: from 2.51% in January to a 2.21–2.40% corridor now, with the network's own rate essentially unchanged at ~2.6% and participation at a record. Yield fell not because staking got less popular but because more people are splitting the same issuance — and this week we can size the other side of that ledger: with the 30-day average base fee at 0.179 gwei against the 13.384 gwei needed to hold supply flat, burned fees come to just 1.3% of issuance (1,160.53 ETH against ~87,700 ETH issued over the trailing 30 days). Effectively none of the network's payout is funded by demand for blockspace; it is new issuance, split more ways.

The rETH parity resolution is worth dwelling on because it retires an argument, not just a number. For three weeks the decentralized tier's yield read was either stale or un-testable, which kept alive the claim that decentralization carries a yield premium or penalty worth trading on. With same-source, same-date reads now confirming a 1–6 basis point difference, the claim is dead at current magnitudes: the market prices stETH and rETH as the same yield instrument, differentiated by liquidity and trust, not by rate. That is bad news for anyone underwriting a "decentralized premium" thesis and good news for this column's simpler model — the corridor's width is set by fee schedules and by what the consensus layer pays, and the user-net band of 2.21–2.40% is the honest summary of both.

7. What to watch into the new window

• Whether the 2.21% or the 2.40% end of the corridor proves correct. A same-basis weekly refresh settles it, and it is the single most useful number this column can nail down.
• Sept 15–16 FOMC — a hike widens the carry gap; a hold re-inflates the trade.
• Whether the entry queue keeps growing at 2.07M ETH. Growth alongside a falling yield is the strongest possible evidence that this is not a yield trade.
• Spark's −21.24% week — the only idiosyncratic move on the DeFi board.
• The validator exit queue. Zero all week. It is the best early-warning indicator for stETH peg stress that exists, it is public, and it moves faster than any yield number on this page. Check it before the corridor.

FAQ

Related reading

• Weekly Intelligence — Week of August 31–September 6, 2026
• The Empty Exit Queue: 2.07M ETH Is Waiting 36 Days to Earn Less Than Cash