Approximately 34% of Ethereum's circulating supply — 41.4 million ETH — is now locked in staking contracts, up from 29% at the start of 2026. Validator rewards have compressed to 2.62% annually, the lowest in three years. At the center of the concentration sits Bitmine Immersion Technologies (NYS...
Approximately 34% of Ethereum's circulating supply — 41.4 million ETH — is now locked in staking contracts, up from 29% at the start of 2026. Validator rewards have compressed to 2.62% annually, the lowest in three years. At the center of the concentration sits Bitmine Immersion Technologies (NYSE: BMNR), a single publicly traded company that holds 5.07 million ETH staked through its proprietary MAVAN validator network, representing roughly 12% of all staked ETH and 4.2% of total circulating supply.
The firm's leveraged financing structure — $11.87 billion raised through at-the-market equity issuance in nine months, plus a $350 million preferred stock offering carrying a 9.5% annual dividend — has created a scenario where a forced liquidation could flood Ethereum's exit queue for weeks. The Kiln incident of September 2025, in which 1.6 million ETH triggered a 46-day exit queue, demonstrated the mechanism. Bitmine's staked position is more than three times larger.
Two competing proposals now address different facets of the problem. EIP-8061, likely included in the Glamsterdam upgrade (Q4 2026), would quadruple the exit churn limit. EIP-8361, introduced August 4 by Ethereum Foundation researcher Justin Drake, would progressively burn validator issuance to zero once 50% of ETH is staked — effectively penalizing large accumulators. Neither addresses both exit mechanics and economic incentives simultaneously.
Ethereum's staking landscape in August 2026 is dominated by a small number of entities. According to on-chain data and public filings:
| Entity | Staked ETH (approx.) | Share of Total Staked | Share of Total Supply | |--------|--------------------:|---------------------:|---------------------:| | Lido Finance | 8.86M | 21.4% | 7.3% | | Bitmine (MAVAN) | 5.07M | 12.2% | 4.2% | | Coinbase | ~5.04M | 12.2% | 4.2% | | Kraken | ~2.5M | ~6.0% | ~2.1% | | Other | ~20.0M | 48.2% | 16.5% | | Total | ~41.4M | 100% | 34.0% |
The top three entities collectively control approximately 45.8% of all staked ETH. Lido's share has declined from 32% in late 2023 to roughly 21.4%, easing one concentration concern. But Bitmine, which did not exist as an ETH staker 18 months ago, has rapidly filled the gap.
Within the liquid staking segment specifically, Lido retains approximately 62% market share, according to Datawallet data. Solo validators and smaller operators continue to be crowded out by institutional-scale participants offering liquid staking tokens (LSTs) or operating proprietary infrastructure like Bitmine's MAVAN network.
The entry queue for new validators stood at 39-50 days as of mid-August 2026, with 2.23 million ETH waiting to begin staking. The exit queue, by contrast, held just two validators — a calm that masks the structural vulnerability underneath.
Bitmine Immersion Technologies, led by market commentator Tom Lee, has executed the most aggressive Ethereum accumulation strategy by any publicly traded entity. The company's SEC filings and earnings reports reveal the scale:
Capital Raised: During the nine months ended May 31, 2026, Bitmine sold approximately 340.7 million common shares through its at-the-market program, raising $11.87 billion after issuance costs. Outstanding common shares increased 149%, from 232.4 million to 579.7 million.
ETH Purchased: Approximately $11.69 billion was spent purchasing ETH over the same period. As of an August 16, 2026 SEC filing, total staked ETH stood at 5,067,309 tokens valued at $9.6 billion.
Revenue vs. Losses: Q3 2026 (quarter ended May 31) showed $46.5 million in staking revenue, with 98% generated through MAVAN. However, the firm simultaneously reported a $92.1 million derivatives loss. Administrative costs surged from $0.74 million to $37.3 million year-over-year. Net loss for the quarter: $83.58 million.
Preferred Stock: In June 2026, Bitmine issued 3.5 million shares of 9.50% Series A Perpetual Preferred Stock at $100 per share, raising $350 million. The preferred pays cumulative weekly cash dividends and ranks senior to common stock. On August 14, the company declared 17 upcoming dividend payments through December 2026.
The financing structure creates a specific risk: if ETH's price declines materially, the company must continue servicing the 9.5% preferred dividend and may face margin calls on derivative positions, potentially forcing ETH liquidations at the worst possible time. The stock price (BMNR) stood at $17.45 as of late August, down 4.58% following the dividend declaration.
Ethereum's validator exit queue is governed by a churn limit — the maximum amount of ETH that can exit staking per epoch (6.4 minutes). Under current rules, this limit is fixed at approximately 256 ETH per epoch, regardless of total staked supply.
When Kiln, then the fifth-largest staking provider with 1.6 million ETH across 51,000 validators, suffered a security breach on September 8, 2025 (hackers drained roughly 193,000 SOL tokens worth $41 million), the company pre-emptively unstaked all its ETH. The results:
According to 21Shares' July 2026 research report, a partial liquidation of approximately 5% of Bitmine's staked ETH (~253,000 ETH, worth roughly $367.5 million at then-current prices) could add an estimated 5-6 days to the exit queue. Under deeper financial distress where the company must liquidate larger portions, the impact scales non-linearly due to the contagion effect observed during the Kiln episode.
Bitmine's staked position of 5.07 million ETH is 3.2x larger than Kiln's was at the time of its incident.
EIP-8061, currently included in the Glamsterdam devnet-1 specification and likely to ship in Q4 2026, addresses the exit-side bottleneck by replacing the fixed churn limit with a formula tied to total staked ETH.
Current limit: ~256 ETH per epoch (fixed) Proposed limit: ~1,258 ETH per epoch at 41.2M ETH staked (approximately 4.9x increase)
According to Figment's institutional staking analysis, if EIP-8061 had been active during the Kiln incident, queue times would have dropped from 46 days to approximately 12 days. For a hypothetical full Bitmine liquidation of 5.07 million ETH, the estimated queue time under EIP-8061 would be roughly 28 days, compared to an estimated 130+ days under current rules.
EIP-8061 addresses exit-side congestion exclusively. There is no corresponding proposal for entry-side queue management on the Glamsterdam roadmap, according to the Ethereum Foundation's Soldøgn interop recap from May 2026.
Sigma Prime's Glamsterdam EIP preference analysis rates EIP-8061 as "should have" status, reflecting consensus among client teams that the current exit throughput is inadequate for a network with 41.4 million ETH staked.
On August 4, 2026, Ethereum Foundation researcher Justin Drake and co-authors introduced EIP-8361, a tapered issuance burn mechanism. The proposal takes a different approach to concentration risk: rather than widening the exit door, it reduces the economic incentive to stake.
Current yield: 2.62% annual at 34% staking ratio Proposed yield at 34%: ~1.2% annual (after 18-month transition) Yield at 50% staking: 0% (issuance fully burned)
The mechanism progressively burns a portion of newly issued ETH as the staking ratio increases, creating a natural economic ceiling on staking participation. At current participation levels, validators would see their yields cut by more than half.
The proposal generated immediate opposition. According to CryptoTimes reporting from August 8, DeFi protocol leaders argued the mechanism would:
By the August 6 All Core Devs consensus call, the recorded next step was for the presenting author to consider withdrawing EIP-8361 from Hegotá (the upgrade following Glamsterdam) consideration. The proposal remains formally open but faces significant community resistance.
For Bitmine specifically, a cut from 2.62% to 1.2% yield would reduce projected annualized staking revenue from approximately $235 million to roughly $108 million — while the 9.5% preferred dividend obligation and operational costs remain fixed.
Three distinct scenarios illustrate the range of outcomes:
Scenario 1: Orderly Operation (Base Case) Bitmine continues staking, ETH price remains stable, and EIP-8061 ships in Glamsterdam. The exit queue capacity quadruples, reducing — but not eliminating — tail risk. Staking yields compress further as participation grows, but Bitmine's scale generates sufficient revenue to service obligations.
Scenario 2: Price-Driven Stress A sustained ETH price decline of 30%+ triggers derivative margin calls (Bitmine reported $92.1M in derivatives losses in Q3 alone). The company must sell ETH on the open market or begin unstaking. At current exit churn limits, unstaking 5.07 million ETH would take an estimated 130+ days. Even under EIP-8061's expanded limits, the queue would stretch to approximately 28 days. Contagion exits, based on the Kiln precedent, could add 25-30% more volume to the queue.
Scenario 3: Yield Compression + Stress If EIP-8361 or a similar proposal is adopted, Bitmine's staking revenue drops to ~$108M while preferred dividend obligations (~$33.25M annually) and operating costs (~$150M annualized based on Q3 run rate) consume the majority. A simultaneous ETH price decline creates a liquidity trap: revenue falls, obligations remain fixed, and exiting requires waiting in a queue that grows longer as others attempt to exit simultaneously.
Ethereum's staking layer has evolved from a decentralized security mechanism into an arena where a single publicly traded company controls more than 12% of all staked ETH, financed by continuous equity dilution and a 9.5% preferred dividend obligation. The exit queue — a rate limiter that processes roughly 256 ETH per epoch — was designed for a network with fewer, smaller stakers. It was not designed for a world in which one entity's financial distress could generate an exit event 3.2 times larger than the Kiln incident, which itself produced a 46-day queue.
EIP-8061 represents a necessary mechanical fix, but quadrupling throughput still leaves the network facing a potential 28-day queue in a Bitmine stress scenario, before accounting for contagion exits. EIP-8361 addresses the incentive layer but risks triggering the very financial stress it seeks to prevent among leveraged stakers. Neither proposal addresses the fundamental question: whether a proof-of-stake network can safely absorb the concentration risk created by corporate treasury strategies borrowed from Bitcoin mining's playbook.
The exit queue's current state — two validators, zero wait — is not evidence of health. It is evidence that no one is stressed yet.