A sudden shift in Ethereum staking is draining billions from exchanges toward a new corporate elite

By the end of 2025, a corner of the market most Ethereum traders rarely watch had built a position large enough to matter for everyone else.

Everstake’s annual Ethereum staking report estimates that public companies’ “digital asset treasuries” collectively held roughly 6.5–7.0 million ETH by December, which is more than 5.5% of the circulating supply.

The number is huge, but the more important part is why these companies chose ETH in the first place.

Bitcoin’s corporate-treasury playbook is built around scarcity and reflexivity: buy coins, let the market re-rate the equity wrapper at a premium, then issue stock to buy more coins.

Ethereum adds a second leg that Bitcoin can’t. Once ETH is acquired, it can be staked, meaning it can earn protocol-native rewards for helping secure the network. Everstake frames that reward stream at roughly 3% APY for treasury-style operators.

A corporate ETH treasury is trying to be a listed vehicle that holds ETH, earns additional ETH through staking, and convinces equity investors to pay for that packaged exposure. The main bet is that the wrapper can compound its underlying holdings over time, and that public markets will finance the growth phase when sentiment is favorable.

The basic mechanics of staking

Ethereum runs on proof-of-stake. Instead of miners competing with computers and electricity, Ethereum uses “validators” that lock ETH as collateral and run software that proposes and attests to blocks.

When validators do the job correctly, they receive rewards paid by the protocol. When they go offline or misbehave, they can lose part of their rewards and, in more severe cases, a portion of the locked ETH through slashing.

Staking is attractive to institutions because the rewards are native to the protocol, not dependent on lending assets to a borrower. It still carries operational risk, but that is dampened by the fact that the core source of yield is the network itself.

Everstake’s report says that by the end of 2025, about 36.08 million ETH was staked, which it describes as 29.3% of supply, with net growth of more than 1.8 million ETH over the year.

That matters for treasuries because it shows staking has become a large, established market rather than a niche activity.

The ETH treasury flywheel: premium financing plus protocol yield

Everstake describes two levers that treasury companies are trying to pull.

The first is mNAV arbitrage. If a company’s stock trades at a premium to the market value of its underlying assets, it can issue new shares and use the proceeds to buy more ETH.

If the premium is large enough, that can increase ETH per share for existing shareholders even after dilution, because investors are effectively paying more for each unit of Ethereum exposure than it costs to acquire ETH directly.

The loop works as long as the premium holds and capital markets stay open.

The second lever is staking rewards. Once the ETH is held, the company can stake it and receive additional ETH over time.

Everstake frames the staking leg as roughly 3% APY, with the key point being low marginal costs once infrastructure is in place. A treasury that stakes wants to compound in token terms, not just ride price appreciation.

Together, the pitch for treasury staking is straightforward. The premium finances growth when markets are optimistic, and staking produces steady accumulation when markets are quieter.

Both mechanisms aim at the same output: more ETH per share.

The three treasury staking playbooks

Everstake’s report concentrates the sector into three large holders and assigns each a role in the story.

It estimates BitMine holds about 4 million ETH, the figure that dominates Everstake’s “hockey stick” chart. Everstake also says BitMine is moving toward staking at an even bigger scale, including plans for its own validator infrastructure and disclosures that “hundreds of thousands of ETH” were staked via third-party infrastructure by late December 2025.

SharpLink Gaming holds about 860,000 ETH, staked as part of an active treasury approach where staking rewards are treated as operating income and remain on the balance sheet.

The Ether Machine holds about 496,000 ETH, with 100% staked. Everstake cites a reported 1,350 ETH in net yield during a period as evidence of what a “fully staked” model looks like.

Those numbers are evidence that the strategy is being institutionalized. These aren’t small experiments for the companies. Their positions are large enough that staking venue, operational posture, disclosure practice, and risk controls become part of the product.

Where institutions stake, and why “compliance staking” exists

The most practical insight in Everstake’s report is that staking is splitting into lanes.

Retail often stakes through exchanges for simplicity, and DeFi-native users chase liquidity and composability through liquid staking tokens.

Institutions often want something closer to traditional operational separation: defined roles, multiple operators, auditability, and a structure that fits existing compliance expectations. Everstake points to Liquid Collective as a compliance-oriented staking solution and uses its liquid staking token LsETH as a proxy for institutional migration.

The report says LsETH grew from about 105,000 ETH to around 300,000 ETH and links that growth to outflows from Coinbase exchange balances as a sign of large holders moving away from exchange custody while still preferring “enterprise-grade” staking structures.

It adds an exchange snapshot that reinforces the point. Everstake says Coinbase’s share fell by roughly 1.5 million staked ETH, from 10.17% to 5.54%, while Binance increased from 2.02 million to 3.14 million ETH, with the share rising from 5.95% to 8.82%.

The figures matter less as a verdict on either venue and more as evidence that staking distribution changes meaningfully when large players reposition.

For treasury companies, that staking-lane question is structural.

If the strategy depends on staking rewards to support compounding, then operator diversification, slashing protection, downtime risk, custody architecture, and reporting practices stop being back-office details and become core parts of the investment case.

The rails underneath the trade: stablecoins and tokenized Treasuries

Everstake doesn’t treat corporate treasuries as a standalone phenomenon, but ties them to Ethereum’s institutional pull in 2025: stablecoin liquidity and tokenized Treasury issuance.

On stablecoins, Everstake says total stablecoin supply across networks surpassed $300 billion, with Ethereum L1 plus L2s holding 61%–62%, or about $184 billion. The argument is that Ethereum’s security and settlement depth keep attracting the on-chain dollar base that institutions actually use.

On tokenized Treasuries, Everstake says the market was approaching $10 billion and puts Ethereum’s ecosystem share at about 57%. It frames Ethereum L1 as a security anchor for major issuers and cites products such as BlackRock’s BUIDL and Franklin Templeton’s tokenized money fund.

This context is important for the treasury trade.

A public company trying to justify a long-term ETH position and a staking program needs a narrative that goes beyond crypto speculation.

Tokenized cash and tokenized Treasuries are easier to defend as structural adoption than most other on-chain categories, and their growth makes it simpler to explain why the asset securing the ledger might matter over a longer horizon.

The risks that can break the Ethereum staking model

Everstake includes a warning about concentration and correlated failures.

It cites a Prysm client outage in December 2025, saying validator participation dropped to around 75% and 248 blocks were missed, and uses the event to argue that client herding can create network-wide fragility.

That risk matters more if large public treasuries consolidate into similar infrastructure choices, because their staking decisions can influence concentration. It also matters because staking returns are only clean when operations are resilient.

While downtime, misconfiguration, and slashing might sound abstract to companies, they are as much part of the business as staking is.

The second risk is capital markets, because mNAV arbitrage is a good mechanism only when markets are strong. If the equity premium compresses, issuing stock becomes dilutive rather than accretive, and the loop stops working.

Staking yield doesn’t fix that on its own, because yield is incremental while equity financing is the growth engine.

A third risk is governance and regulation.

Treasury companies operate inside disclosure and custody regimes that can tighten quickly. The strategy depends on maintaining a structure that auditors, boards, and regulators can tolerate, especially if staking becomes a material contributor to reported income.

The ETH treasury trade is built on a simple proposition: accumulate ETH, stake it to grow holdings in token terms, and use public-market access to scale faster than a private balance sheet could.

Whether it survives as a durable category will depend on two measurable things: how well these companies operationalize staking without creating hidden fragility, and how consistently their equity wrappers can hold premiums that make the financing loop work.