Nominal yield is the least useful number in this report. The useful parts are utilisation, the source of the return, and the spread over a real reference rate.
Every level below is recomputed for the week ending 6 September 2026 from per-pool historical data, so the comparison column genuinely reflects the prior week rather than a later snapshot. Rate moves of a few basis points should be read as noise; anything beyond roughly half a point is worth explaining, and one row this week clears that bar.
Quick answer
Where were DeFi yields highest in the week ending 6 September 2026? Arbitrum offered the highest stablecoin lending rate among tracked chains at 4.87% TVL-weighted, ahead of Base at 4.51% and Ethereum at 4.09%. Solana staking, at 4.41%, sat below its level a week earlier after the week's largest move. Judged as a spread over the tokenized T-bill reference of 3.33%, Arbitrum paid the widest premium.
1 Yield dashboard
2 Week-over-week comparison
| Aggregate | Week ending 2026-08-30 | Week ending 2026-09-06 | Change | Note |
|---|---|---|---|---|
| Ethereum stablecoin lending | 4.00% | 4.09% | +0.09pp | Deepest pool set in the sample |
| Tokenized T-bill funds | 3.32% | 3.33% | +0.01pp | The reference rate every other row is judged against |
| ETH liquid staking | 2.21% | 2.21% | +0.00pp | Pure staking venues; lending collateral excluded |
| Solana staking | 4.97% | 4.41% | -0.56pp | Includes MEV and priority-fee components |
| Blended stablecoin, all chains | 3.98% | 4.01% | +0.03pp | TVL-weighted across tracked chains |
| Base stablecoin lending | 4.61% | 4.51% | -0.10pp | Smaller pool set |
| Arbitrum stablecoin lending | 4.79% | 4.87% | +0.08pp | Smaller pool set |
| BSC stablecoin lending | 3.05% | 3.02% | -0.03pp | Smaller pool set |
| Tron stablecoin lending | 3.70% | 3.71% | +0.01pp | Smaller pool set |
Read the change column with a tolerance of roughly ten basis points. Inside that band you are mostly looking at sampling noise: which pools were included, when each venue reported, and how the weighting moved as TVL shifted between pools.
Solana staking is the only row that clearly exceeds it. It is also the only row with a plausible mechanism attached, given the network's issuance change a week earlier — and the only one worth putting on a watchlist rather than merely recording.
3 Spread over the risk-free reference
| Strategy | Yield | Reference risk-free | Spread |
|---|---|---|---|
| Tokenized T-bills | 3.33% | 3.33% | 0.00% |
| Ethereum stablecoin lending | 4.09% | 3.33% | 0.76% |
| Arbitrum stablecoin lending | 4.87% | 3.33% | 1.54% |
| Base stablecoin lending | 4.51% | 3.33% | 1.18% |
| Solana stablecoin lending | 3.84% | 3.33% | 0.51% |
| BSC stablecoin lending | 3.02% | 3.33% | -0.31% |
| TRON stablecoin lending | 3.71% | 3.33% | 0.38% |
| Solana LST | 4.41% | 3.33% | 1.08% |
| ETH LST | 2.21% | 3.33% | -1.12% |
Judged as spreads, the ordering changes from the one nominal rates suggest. Arbitrum pays the widest premium and stays there for the same reason it did last week: a smaller pool set where rates have not yet been arbitraged toward the reference.
The row that should give an allocator pause is ETH liquid staking, which pays less than the T-bill reference. That is not a flaw — staking is not a lending strategy, and its return includes token exposure rather than a credit premium. It does mean that choosing it on yield alone misreads what you are buying.
4 Lending rates by market
| Venue | Chain | Asset | APY | Pool size |
|---|---|---|---|---|
| sky-lending | Ethereum | SUSDS | 3.6% | $4.7B |
| maple | Ethereum | USDC | 5.0% | $2.7B |
| circle-usyc | BSC | USYC | 3.0% | $2.6B |
| ethena-usde | Ethereum | SUSDE | 4.6% | $1.4B |
| ondo-yield-assets | Ethereum | USDY | 3.6% | $1.2B |
| maple | Ethereum | USDT | 4.7% | $1.0B |
| blackrock-buidl | Solana | BUIDL | 3.6% | $1.0B |
| centrifuge-protocol | Ethereum | USDS | 3.2% | $0.8B |
| blackrock-buidl | Ethereum | BUIDL | 3.6% | $0.7B |
| invesco-ustb | Ethereum | USTB | 3.3% | $0.6B |
The spread between venues inside a single chain remains wider than the spread between chains, which is the practical argument for reading pool-level rates rather than chain averages. A composite will always look like something nobody can actually deposit into.
5 Liquid staking yields
| Asset | Chain | Yield | Issuer | TVL |
|---|---|---|---|---|
| STETH | Ethereum | 2.2% | lido | $24.2B |
| WEETH | Ethereum | 2.3% | ether.fi-stake | $5.3B |
| RETH | Ethereum | 2.2% | rocket-pool | $3.4B |
| RSETH | Ethereum | 2.2% | kelp | $1.1B |
| CBETH | Ethereum | 2.3% | coinbase-wrapped-staked-eth | $0.5B |
| OSETH | Ethereum | 2.2% | stakewise-v3 | $0.4B |
| WSTETH | Ethereum | 0.0% | fluid-lending | $0.2B |
| ETH-STETH | Ethereum | 1.2% | curve-dex | $0.1B |
Lido remains the dominant route by size, and its rate sits close to the sample average. Where providers differ by more than about twenty basis points, the gap is usually attributable to how much of the yield comes from MEV rather than from protocol issuance — which is also the component most likely to move without warning.
6 Stablecoin yields
| Chain | TVL-weighted APY | Stablecoin TVL | Pools counted |
|---|---|---|---|
| Arbitrum | 4.9% | $1.4B | 12 |
| Base | 4.5% | $2.3B | 12 |
| Ethereum | 4.1% | $15.1B | 13 |
| Solana | 3.8% | $1.9B | 12 |
| Tron | 3.7% | $0.5B | 3 |
| BSC | 3.0% | $3.1B | 12 |
Two things determine these levels, and only one is durable. The durable part is underlying real-world rate exposure from tokenized Treasuries. The cyclical part is perp funding, which feeds several of the higher-quoting products and can reverse in a day without anything visibly changing in the pool.
7 Playbook: matching strategy to mandate
| If your goal is | Consider | Accept | Avoid |
|---|---|---|---|
| Cash preservation | Tokenized T-bill funds | Issuer, custody and redemption-timing risk | Treating a redemption window as instant liquidity |
| Working capital | Blue-chip lending on Ethereum | Smart-contract and governance risk | Chasing an isolated market's premium without reading its collateral |
| Income with a risk budget | Isolated or smaller-chain lending | Collateral and chain concentration | Assuming the premium compensates for concentration by default |
| Tactical carry | Funding-linked stablecoin products | A funding regime that can flip for days | Treating a cyclical rate as structural |
| Long-horizon exposure | Liquid staking | Token price risk alongside the yield | Comparing its rate to a credit spread |
8 Formulas used
9 What we would watch next week
- Whether Solana staking stabilises or falls again — two consecutive moves would warrant its own analysis rather than a footnote.
- Whether the Ethereum lending premium over T-bills holds above three-quarters of a point, which has not been the case for most of the six-week window.
- Whether Arbitrum's premium persists as its pool set grows, which would suggest the gap is structural rather than a size effect.
- Whether anything this week changes how much blended stablecoin yield traces to real-world rates versus trading activity.
Key Takeaways
- Almost every level held; only Solana staking moved materially, down 0.56pp.
- Ethereum lending pays 0.76pp over tokenized T-bills, up from 0.68pp.
- Arbitrum stablecoin lending offers the widest spread among tracked chains — 4.87% APY, 1.54pp over the reference.
- ETH liquid staking still pays less than the risk-free reference — a different product, not a worse one.
- This edition's comparison column reflects the prior week's data, not a later snapshot.
10 In brief
Almost nothing moved, apart from one line. Ethereum stablecoin lending rose nine basis points to 4.09%, the tokenized T-bill reference gained a basis point to 3.33%, and Ethereum liquid staking was unchanged at 2.21%.
Solana staking was the exception, falling from 4.97% to 4.41% — the largest single move anywhere in our sample this week. Staking yield carries a MEV and priority-fee component that varies week to week, so one reading does not establish a trend. It does sit in the direction implied by Solana's governance decision the week before, which doubled the network's disinflation rate.
The spread story is intact and slightly stronger. Ethereum stablecoin lending pays 0.76pp over the tokenized T-bill reference, up from 0.68pp a week earlier. Anyone deciding between the two is being paid more this week to accept protocol risk, not less.
Sources & Methodology
- DefiLlama yields endpoint — pool-level APY and TVL, aggregated TVL-weighted.
- Per-week archive built from each pool's own historical chart, so this edition's levels belong to the week named rather than to the day of publication.
Headlines and figures on this page are drawn from the outlets listed above; commentary is clearly labelled opinion and is not investment advice. Last reviewed 2026-09-06.