KoinlyticsBitMine holds roughly 5.777 million ETH. That is close to 4.8% of all ether in existence. About 85% of the position is staked, generating an estimated $247 million in annual rewards at current rates.
To put the size in context: total staked ETH sits above 41 million, a staking ratio of 34.23%. BitMine alone accounts for roughly 12% of everything staked.
The Position Is a Yield Trade, Not a Price Trade
A company holding 4.8% of an asset's supply and staking 85% of it is not positioned for a trade. It is positioned for a coupon.
The arithmetic explains the structure. At roughly $1,900 per ETH, the position is worth around $11 billion. A $247 million annual return on that is roughly 2.2% after the unstaked sleeve is accounted for, which is below the current Treasury yield but comes with an asset that can appreciate. That is a fundamentally different bet than holding Bitcoin, which offers the appreciation without the coupon.
It is the same logic that showed up in Intesa Sanpaolo's Q2 filing, which cut its Bitcoin ETF position 93.7% while nearly tripling a staked Ethereum holding. Institutions comparing a zero-coupon asset against a yielding one in a rate environment like this one keep reaching the same conclusion.
What 85% Staked Actually Means
Staked ETH is not liquid ETH. Exiting a validator requires entering a withdrawal queue whose length depends on how many others are leaving at the same time. Under calm conditions it clears in days. Under stress it does not, and stress is precisely when everyone wants out simultaneously.
For a position this size the practical consequences are:
- Roughly 4.9 million ETH cannot be sold on short notice regardless of price
- The 15% unstaked sleeve, around 870,000 ETH, is the only fast-liquidation capacity
- Any decision to unwind at scale would itself lengthen the queue it has to pass through
- Slashing risk applies to principal, not just to forgone yield, though it remains remote
The Concentration Question
A single entity controlling 12% of staked ETH raises a governance concern that is separate from the financial one. Ethereum's security model assumes validators are numerous and independently controlled. Concentration erodes that assumption even when the concentrated holder is entirely well behaved, because the network's resistance to coordinated action depends on there being nobody in a position to coordinate.
The standard mitigation is distributing across many operators and clients, which large holders generally do. That reduces technical correlation. It does not change who ultimately decides what those validators do.
The staking ratio going up is usually reported as a health metric. It is also a concentration metric, and the two readings point in opposite directions.
The Supply Side Effect
41 million ETH locked in validators is float removed from the market, and float removal cuts both ways. Thinner order books amplify rallies and amplify drawdowns equally. The effective supply available to absorb selling is materially smaller than the circulating figure implies.
This is part of why Ethereum ETF flows have been reversing. Spot ether funds have leaked all year while staked variants gained, and five consecutive positive weeks culminating in $244.9 million for the week ending August 7 reflects money arriving through the yielding wrapper rather than the plain one.
What to Watch
- Whether BitMine's holding grows further, and at what percentage of supply the concentration becomes a governance discussion rather than a market one
- Validator exit queue length during any sustained drawdown, the untested part of this structure
- The staking ratio crossing 35%, which mechanically lowers per-validator issuance
- Whether other corporate treasuries adopt the staked-ETH model over Bitcoin accumulation
- Operator and client diversity within large staked positions, the only real mitigation available
If you hold ETH through a mix of spot, liquid staking tokens and ETF shares, those three have different exit paths, different effective yields after fees and different counterparties. They look like one line on a portfolio screen and behave like three separate instruments the moment you need to unwind, which is exactly when finding out is most expensive.