Weekly DeFi Intelligence — 24–30 August 2026

1 Networks — how active is each chain?

Ethereum remains the settlement and collateral centre of gravity: it carries the deepest liquidity and the largest stablecoin base, and its fee generation reflects that rather than raw transaction count. Its turnover is low, which is what you would expect from capital that is parked as collateral rather than actively traded.

Solana is the inverse: the highest turnover in the sample, with volume driven by short-lived, high-frequency flow. Base sits between the two — solid turnover on a modest TVL base, the profile of a chain whose activity comes from many small consumer transactions rather than large collateral positions.

The practical read is that these chains are not really competing for the same activity. Treating them as interchangeable because they all host DeFi protocols is the most common analytical mistake in this space.

It is worth being explicit about what would change this read. If Ethereum's turnover rose materially while its TVL stayed flat, that would suggest collateral is being put to work rather than parked — a genuinely different regime. If Base's turnover converged down toward Ethereum's, the consumer-activity story would weaken. Neither happened this week.

2 Stablecoins — how are stablecoins performing?

Aggregate stablecoin supply across the chains we track stands at $284.2B, up +0.4% week over week. Ethereum contributed the largest absolute increase (+$0.9B), with TRON's USDT float second (+$0.5B) — and TRON's steady mint cadence has now persisted for several consecutive weeks.

Two things follow. First, more float chasing the same borrow demand is precisely why lending yields compressed this week. Second, concentration is rising: the top two chains hold the clear majority of supply, which is efficient but means a single-issuer disruption would propagate widely.

We also continue to see the market split into payment float and yield-bearing savings instruments. Those categories behave very differently under stress, and blurring them leads to bad conclusions about risk.

One measurement caveat matters here. Stablecoin supply is usually counted where the token is issued or where it settles, and bridged representations can complicate the picture. We treat week-over-week change as reliable and cross-chain levels as approximate, which is the honest way to read the table.

3 DeFi — what happened this week?

Aggregate TVL reached $73.0B, +1.0% week over week, with Solana carrying the move at +6.3% while BNB Chain slipped -1.1%. Concentrated gains like this usually mean rotation or a chain-specific catalyst rather than ecosystem-wide accumulation, so we read the breadth — not the headline — as the signal.

Our anomaly screen — the deviation of each chain-metric pair from its trailing four-week baseline, in standard deviations — flags six pairs this week. The largest sit on Arbitrum: stablecoin supply (+3.3%, 7.8σ above baseline) and fees (+100%, 4.5σ), with Solana's fees (+18.8%, 2.7σ) also elevated.

The most structurally interesting development remains tokenized Treasuries crossing a reference AUM threshold. It changes what a conservative on-chain allocation can look like, and it increasingly competes with lending as the default home for idle stablecoin balances.

On the risk side, the week's losses came from collateral pricing rather than contract bugs: a manipulated reUSD quote hit a Morpho market, Moonwell's Base markets froze after an inflated collateral quote, and Tectonic on Cronos lent against a thinly traded token. The modal loss vector this week was a risk-parameter decision on thin collateral, not user-side approval hygiene.

Two caveats keep us honest about TVL. It can double-count recursive strategies, and it says nothing about whether the capital is doing anything useful. That is why we pair every TVL figure with turnover and fees rather than quoting it alone.

4 Yields — what is happening in the yield market?

The levels first: stablecoin lending pays between 3.4% and 4.7% across the large chains, with Ethereum — the deepest market — at 4.23%. L2 venues pay a premium over Ethereum, which is compensation for thinner liquidity rather than free return. More float chasing the same borrow demand puts downward pressure on these rates; the direction of that pressure, not any single week's move, is what we track.

Liquid staking pays 2.20% on Ethereum and 4.55% on Solana — the highest staking yield in our sample, though nominal yield is not the same as risk-adjusted return — validator-set and token risk both need to be priced in.

Blended across every tracked chain, stablecoin strategies pay 4.18%. The blended number hides wide dispersion between chains, which is precisely why we publish the per-chain table rather than a single headline rate.

The framing that matters most is the spread. With tokenized T-bill funds — the on-chain proxy for the risk-free rate — yielding around 3.47% after fees, the reference point for any strategy is roughly 3.5%. This week that reference sits below Ethereum stablecoin lending at 4.23%, so blue-chip lending is being paid a genuine premium over the on-chain risk-free rate — unusual, and worth watching for how long it lasts.

5 Outlook — what to watch next week

This is a watchlist, not a forecast. Each row has an explicit trigger so that watching is a defined activity rather than a vague intention. If none of the triggers fire, the correct conclusion is that nothing changed.

6 What would change our mind

We publish falsifiers because a view that cannot be contradicted is not analysis. If the evidence in the right column appears, we will say so rather than explaining it away.

7 Risk register

We keep this register standing rather than rewriting it weekly, because risks that persist are usually more important than risks that appear. Only the likelihood column tends to move.

8 Formulas used in this briefing

9 The numbers behind the five answers

We publish this table because a briefing that states conclusions without the underlying figures is unfalsifiable. If you disagree with an answer above, the metric behind it is here to argue with.

10 How to read this briefing

1. Skim the scorecard for direction; read only the sections whose direction changed.
2. Use the numbers table to check any claim you intend to act on.
3. Read the falsifiers before the outlook — they tell you how confident we are.
4. Treat the risk register as standing context rather than weekly news.

11 In brief

Three of the five questions have the same answer as last week, and that stability is the headline. TVL and stablecoin supply both grew while trading volume cooled, and stablecoin lending held above staking across the large chains. The one genuinely developing story is the growing role of tokenized real-world assets in setting a conservative yield floor.

We would rather report a quiet week than manufacture drama. Stability in the data is information, and it argues against repositioning on noise.

12 Weekly scorecard