KoinlyticsEthereum's staking ratio has reached 34.23%. More than 41 million ETH now sits in validator nodes, a staking market capitalisation above $77 billion at ETH's current $1,880. Roughly one in three ether in existence is locked in consensus rather than circulating.
That number has been climbing steadily all year and it is now doing something it did not do in previous cycles: shaping institutional allocation decisions in a way that shows up directly in filings.
The Filing That Made It Concrete
Intesa Sanpaolo, Italy's largest bank, disclosed a Q2 2026 13F that cut its BlackRock iShares Bitcoin Trust position by 93.7%, from 646,809 shares to 40,723, while nearly tripling its iShares Staked Ethereum Trust holding from 116,200 to 349,600 shares. It also cut bullish IBIT calls by 99.3% and opened a put referencing 500,000 IBIT shares.
The bank did not leave crypto. It moved from an instrument that pays nothing to one that pays roughly 2.8% to 3.5% annually. For a treasury desk comparing a zero-coupon asset against a yielding one in a year where the underlying has been range-bound and Treasury yields sit at multi-decade highs, that comparison resolves quickly.
Where the Yield Comes From
Staking rewards are not a promotional rate or a subsidy. They come from two sources: protocol issuance paid to validators for securing consensus, and priority fees paid by users for transaction inclusion. A staked ETH ETF holds ether in validator nodes, collects both, deducts the operator cut and the fund expense ratio, and distributes the remainder.
That structure has three consequences worth understanding before treating the yield as a bond coupon:
- The rate is variable. Issuance falls as the staking ratio rises, so a higher participation rate mechanically dilutes per-validator returns. At 34.23% and climbing, the yield curve points down.
- Priority fee income scales with network activity. Quiet months pay less.
- Slashing is a real if remote risk. Validator misbehaviour or infrastructure failure can burn principal, not just forgo yield.
The Liquidity Question
41 million ETH locked in validators is not 41 million ETH available to sell. Exiting a validator requires entering a withdrawal queue whose length depends on how many others are exiting simultaneously. Under calm conditions the queue clears in days. Under stress, when everyone wants out at once, it does not.
This is the risk that is easiest to ignore during a range-bound market and hardest to manage during a fast one. An ETF wrapper adds a layer: the fund can meet redemptions from its cash buffer and unstaked sleeve without touching the validator set, but only up to the size of that buffer. Beyond it, the queue applies to the fund exactly as it applies to anyone else.
For a bank holding 349,600 shares that is an acceptable operational detail. For a market where a third of supply is behind a queue during a genuine liquidation event, it is a structural feature that has not yet been tested at scale.
The Supply Side Effect
A rising staking ratio removes float. Fewer liquid ETH means thinner order books, which cuts both ways: it amplifies rallies and it amplifies drawdowns. It also means the effective supply available to absorb selling pressure is materially smaller than the headline circulating figure suggests.
Spot ETH ETFs have had a mixed year, with $365 million of July inflows reversing $528 million of June and $540 million of May outflows, and a $12.2 million outflow on August 4 against $170.3 million into Bitcoin ETFs. But looking only at plain spot ETH products understates institutional Ethereum demand, because a growing share of that demand is arriving through staked wrappers that get counted separately or not at all.
Bitcoin's institutional pitch is that it goes up. Ethereum's is increasingly that it pays you while you wait. In a year where nothing has gone up much, the second pitch has been landing better.
What to Watch
- Staking ratio crossing 35%, and the corresponding drop in per-validator issuance
- Combined staked plus spot ETH ETF flows as a single figure, the only version that reflects real demand
- Validator exit queue length during any sustained drawdown, the untested part of the structure
- Pass-through rate competition between issuers, currently 2.8% to 3.5% and likely to compress
- Whether other European and Asian bank filings show the same Bitcoin-to-staked-Ethereum rotation
If you hold staked ETH directly, through a liquid staking token, or through an ETF wrapper, the three have different exit paths and different effective yields after fees. A blended position that looks like one allocation on a portfolio screen can behave like three separate instruments the moment you need to unwind it.