Bitmine Ethereum Treasury Hits 4.97M ETH as Tom Lee Defends Staking Bet

The Bitmine Ethereum treasury just got heavier. Last week the company, trading as BMNR, added more than 100,000 ether to its pile, roughly $230 million at current prices, and it now sits on 4,976,485 ETH worth about $11.9 billion as of Monday, April 20. That is not a position. That is a policy. And the person writing the policy is chairman Tom Lee, who has turned a former bitcoin miner into the largest staked-ether entity on the network.

How the Bitmine Ethereum Treasury Got So Big So Fast

Rewind to June 30, 2025. Bitmine Immersion Technologies was still pitching itself as an immersion-cooled bitcoin mining outfit, a business that had already cycled through names like Sandy Springs Holdings and Renewable Energy Solution Systems since 2019. That morning the company announced a $250 million private placement and pivoted the whole business toward stacking ether. The stock more than doubled from open to close on a single day, even though mining ETH has been physically impossible since The Merge in 2022.

Ten months later, the accumulation is staggering. According to the company's own Bitmine Ethereum treasury disclosure filed with the SEC, Bitmine's crypto and cash holdings now total $12.9 billion. Of that, 3,334,637 tokens are locked up validating the network. The rest sits liquid, ready to deploy on dips. It is the kind of concentrated bet you usually only see from sovereign wealth funds, and Lee is running it from a company most retail investors could not have named a year ago.

Bitmine has staked more ETH than other entities in the world. At scale, the projected ETH staking reward is $330 million annually. Annualized staking revenues are now $221 million.

Why Tom Lee Keeps Calling Ether a Wartime Store of Value

Lee does not dress up the thesis. "ETH is the wartime store of value," he said in April 2026, a line he has repeated in interviews and on the company's official Tom Lee chairman page. The framing is deliberate. Bitcoin gets the peacetime narrative, the digital gold pitch. Lee is arguing that programmable money with native yield and a burn mechanism behaves differently when the world gets messy, because it generates income while it sits.

That income is the whole point. Lee is not running a momentum trade. He is running a cash-flow business dressed up as a crypto treasury. Every staked token pays a coupon. Every coupon compounds. And every quarter, the gap between passive bitcoin treasuries and Bitmine's model gets wider, at least on paper. Whether the market rewards that distinction with a premium valuation is the open question, and it is the one short sellers will press hardest if ether stays below water.

How Does Ethereum Staking Actually Generate That Yield?

The short answer: proof-of-stake pays validators to secure the chain.

Mining ether ended on September 15, 2022. The Merge swapped proof-of-work for proof-of-stake and cut daily issuance by about 88%. What replaced the mining reward is something bitcoin has never had: a built-in yield paid to anyone willing to lock tokens and run a validator. According to the Ethereum Foundation's own Ethereum staking documentation, validators currently earn between 2.8% and 3.4% annually, and more than 36 million ETH, roughly 30% of the total supply, is already committed to staking contracts.

For Bitmine, the arithmetic is simple and brutal. Lock 3.3 million tokens, earn a mid-single-digit yield in ether, and the treasury grows in the denomination that matters without issuing a single new share. That is how a company with no mining operation and no customer product generates $221 million of annualized revenue. The network pays them to hold.

• Validators lock at least 32 ETH to activate a node, though institutions pool far more.
• Rewards accrue in ether, not fiat, so the income scales with price.
• Slashing penalties apply if a validator misbehaves or goes offline for too long.
• Withdrawals have been live since the April 2023 Shapella upgrade.

The Burn Mechanism Nobody Outside Crypto Talks About

EIP-1559 changed the plumbing in August 2021. Every transaction on Ethereum pays a base fee, and that base fee gets destroyed. Permanently. Not redirected to validators. Burned. Combine that with the post-Merge collapse in issuance, and you get a monetary system where supply responds to demand in real time. Busy network, high fees, heavier burn, tighter supply. Quiet network, low fees, modest burn, mild inflation. It is a feedback loop written into code.

Right now the loop is running in mild-inflation mode. Ether supply sits near 120 million tokens with annual inflation around 0.23% as of April 2026, a function of layer-2 upgrades in 2024 and 2025 that drove base fees through the floor. For comparison, gold's new annual supply runs near 2%, according to World Gold Council data. Ethereum's inflation rate is roughly one-ninth of gold's, and that gap widens every time network activity picks up and the burn rate spikes.

Call this the part of the story the original piece got right and most mainstream coverage still ignores. The asset argues against its own inflation. No other major monetary asset in history has that property wired directly into its protocol.

Self-Repaying Loans, Paper Losses, and the Part That Should Worry You

Here is where the elegant-strategy framing starts to wobble. Sophisticated investors have been combining staking yield with on-chain borrowing to create what DeFi veterans call a self-repaying loan. You post staked ether as collateral, borrow against it, and let the staking yield grind down the loan balance. The smart contract executes without a bank, without a lawyer, without anyone's permission. Done right, the asset retires its own debt while you keep the upside.

Bitmine, to be clear, has not publicly disclosed any such use on its own stack. The company's position is that it simply accumulates, stakes, and holds. That is the honest read of the filings. But the broader pitch floating around finance Twitter, the one that says anyone can do what the ultra-high-net-worth crowd figured out 18 months ago, deserves more scrutiny than it is getting. Liquidation risk is real. Smart-contract risk is real. And ether trading well below its August 2025 high of $4,950 means anyone who borrowed near the top is not sitting on an elegant wealth structure, they are sitting on a margin call waiting to happen.

Bitmine itself is nursing paper losses in the billions against its average acquisition cost. Lee calls it a buying opportunity. Shareholders who came in near the top of the run are calling it something else. The truth is somewhere in the middle: the math on what Ethereum is as a network has not changed, but the math on when you bought matters a lot.

You can agree or disagree with his view. But the math on what Ethereum actually is does not change because the price went down.

What This Means for the Next Wave of Crypto Treasury Companies

Bitmine is now the template nobody wanted to admit existed. MicroStrategy built the bitcoin version. Bitmine is building the ether version, except with a native yield attached. If the model holds through a full cycle, expect a parade of copycats filing S-1s and rebranding their treasuries accordingly. If it breaks, expect the same short sellers who circled MicroStrategy in earlier cycles to sharpen their knives.

The tell will be in the next two quarters. If Bitmine keeps reporting rising staking revenue and resists the temptation to lever up the balance sheet, the thesis holds. If it starts borrowing against its stash to buy more, the self-repaying loan becomes a self-inflicted wound the moment ether has a bad month. Lee has not taken that step yet. Watch whether he can resist it.