The 66% Illusion: Reading Bitcoin's Ownership Ledger When the Numbers Don't Vote Together

Hook

Silence is the first vote in a true consensus.

I have spent the better part of a decade auditing ledgers, and I have learned that the loudest claims are usually the least examined. So when an ownership report crossed my desk last week — one promising to answer who really owns bitcoin, framed as a contest between Wall Street and ordinary people — I did what I always do. I ignored the headline and went straight to the arithmetic.

The headline offered a tidy story: ordinary people hold 66% of all bitcoin, roughly 13.9 million coins, while institutions claw at the remainder. It is a comforting story. It is also a story that, on closer inspection, does not add up.

The number that stopped me was not the 66%. It was 696,000. According to River's data, individual holders net sold 696,000 bitcoin across 2025. Over the same window, corporations, funds, and governments net bought close to one million. That leaves a gap of roughly 300,000 coins that no seller's ledger can account for. In a system with a hard cap and a public ledger, a 300,000-coin hole is not a rounding error. It is a confession that our categories are wrong — and that the map we use to describe ownership no longer matches the territory it claims to chart.

Context

Bitcoin's supply has never been static, but for most of its sixteen-year life the movement was glacial: coins mined, coins lost, coins hoarded. That changed in January 2024, when the United States approved spot bitcoin ETFs and the asset was formally inducted into the machinery of traditional finance. The induction was not merely symbolic. It created a new class of holder — the custodial fund — whose behavior is governed not by conviction but by inflows, redemptions, and the calendar of institutional allocation.

The distribution, as reported by aggregation platforms such as Bitbo and River, reads like this. Individuals hold 13.9 million coins, or 66% of the eventual 21 million cap. ETFs hold 1.48 million, about 7%. Corporations hold 1.66 million, about 8%, of which Strategy alone accounts for 848,000. Governments hold 518,500, roughly 2.5%, most of it seized in enforcement actions. An estimated 1.6 million coins are permanently lost — keys discarded, drives destroyed. Satoshi's 968,000 coins, about 4.6%, have never moved. Together these identified categories cover roughly 96% of supply, leaving about one million coins unaccounted for.

Read plainly, those numbers support a happy narrative: the little guy still owns most of the pie. But I have audited too many balance sheets to trust a category label without opening it up. And when you open up "individual," the 66% begins to dissolve.

I learned this discipline the hard way. In 2017, as a senior researcher at a Tallinn cybersecurity firm, I spent four months reconstructing the The DAO hack from Etherscan logs, cataloguing fourteen distinct flaws in the reentrancy logic. The lesson was not the vulnerability itself. It was that the labels people attached to the system — decentralized, autonomous, trustless — described an aspiration, not a fact. I wrote a thirty-page paper afterward titled Code is Not Law: The Moral Vacuum in Smart Contracts, arguing that technical efficiency without ethical governance produces harm at scale. The same instinct applies here. A category that flatters our politics deserves more scrutiny, not less.

Core Insight

The "individual" bucket, on inspection, is a container into which at least three incompatible things have been poured.

First, it almost certainly includes coins that no living person can spend. The 1.6 million lost coins are still on the ledger, still counted in every supply chart, but they are votes that can never be cast. Second, it includes Satoshi's 968,000 dormant coins, which have not moved in fifteen years and are held — if held is even the word — by a ghost. Third, and most consequentially, it includes exchange custodial wallets. When a retail user buys bitcoin on a major exchange, the coins sit in a pooled hot or cold wallet controlled by that exchange. On-chain, that wallet is a single large address. In the statistics, it may be counted as individual — yet economically, those coins are intermediated by a corporation with its own solvency risk. After FTX, no one should need convincing that custodial coins are not the same as self-custodied coins.

Subtract those three categories and the picture inverts. The genuine free float — coins that are self-custodied, liquid, and spendable by individuals — plausibly sits between 10 and 12 million, not 13.9 million. That is a meaningful correction. The claim that ordinary people own two-thirds of bitcoin survives only if we count the dead, the ghost, and the custodians as ordinary people.

Silence is the first vote in a true consensus, and the ledger has been remarkably silent about its own accounting.

The more interesting story, however, is not on the individual side. It is on the institutional side, where a genuinely novel structure has emerged and where, I believe, the real fragility of this cycle resides: the corporate treasury model.

Strategy — formerly MicroStrategy — is the archetype. It holds 848,000 coins, more than half of all corporate holdings. The model is elegant on paper. The company issues equity or convertible debt at a premium to its bitcoin net asset value, uses the proceeds to buy more bitcoin, and thereby pushes the net asset value higher, which sustains the premium, which permits more issuance. It is a flywheel, and like all flywheels it depends on a continuous supply of momentum. When the premium is wide, the machine prints. When the premium narrows, the machine stalls.

There are early signs of stalling. In a single week in early October, Strategy spent $142.5 million servicing interest and dividends and another $73.7 million repurchasing its own preferred stock — while buying only 334 bitcoin, at a cost of roughly $28.7 million. Read that ratio slowly. The company spent five times more maintaining its capital structure than it spent acquiring the asset that justifies the structure. That is not a growth posture. That is a maintenance posture. And maintenance, in a model built on momentum, is the first chapter of a different story.

The preferred-stock buyback is itself a signal. Preferred equity is the expensive, last-resort capital. When a company repurchases it while slowing its core purchases, the most charitable reading is prudent balance-sheet management. The less charitable reading — the one an auditor learns to prefer until proven otherwise — is that the ordinary-equity issuance window is narrowing, and the company is being pushed toward costlier instruments to keep the flywheel turning.

The metric to watch is the mNAV: the ratio of market capitalization to the net asset value of the coins held. As long as mNAV stays meaningfully above one, the issuance engine works. If it converges toward one, the arbitrage that funds the purchases disappears, and the model's expansion phase ends. It is not a question of whether the company believes in bitcoin. It is a question of arithmetic.

Then there is Metaplanet, the Japanese imitator, which holds 44,000 coins — about 2.6% of the corporate sector — and which allocated 85% to 90% of its assets to bitcoin. In the third quarter it sold 10,000 coins and repurchased 11,000, a maneuver it described as demonstrating liquidity. I have heard that language before. During the 2020 DeFi summer, when I consulted on governance tokenomics for a mid-sized DAO, I watched teams use similar phrases — demonstrating liquidity, optimizing treasury — to dress up what were, functionally, distress signals. I helped that project adopt quadratic voting to blunt whale dominance, and the real lesson was not algorithmic. It was that honest governance begins with honest language. A company does not need to demonstrate liquidity when it has liquidity. The demonstration is the tell.

Here I should be precise about what the data can and cannot support. The ownership figures all flow from third-party aggregators — Bitbo, River — whose methodologies are not fully public. Do they include exchange wallets? Do they exclude provably lost coins? Do they net out coins already counted inside ETF custody? We do not know, and that opacity matters. There is also a troubling internal inconsistency in the source material: the report cites a bitcoin price near $83,000, yet the holdings it describes — 848,000 coins at Strategy, tens of thousands at Metaplanet — correspond to a later, higher-priced era. Price and holdings appear to be stitched from different moments. When a dataset cannot agree with itself about time, it cannot be trusted about magnitude. Treat the percentages as directional, not definitive.

What the directional picture does tell us is that the marginal buyer of bitcoin has changed identity. The individual is no longer accumulating; on the available evidence, the individual is selling — 696,000 coins in a single year. The replacement buyer is institutional: ETFs, corporate treasuries, and, more passively, governments holding seized assets. This is what market technicians call a transfer from weak hands to strong hands. It is generally bullish for scarcity. But it introduces a dependency that did not previously exist.

Follow the chain one more step, because the transmission matters. Miners are the only structurally forced sellers in the system: they must convert coins to cover energy and hardware regardless of price. If corporate treasuries and ETFs stop absorbing new supply, the miner sell flow becomes the dominant marginal pressure, and the price must fall until some other buyer clears it. The whole architecture currently rests on the assumption that institutional demand is elastic and persistent. Nothing in the data guarantees that. ETF flows are bidirectional by design — a single redemption notice can turn this cycle's most reliable bid into its most reliable offer. History already shows single-week outflows in the billions. A strong hand that can be asked to leave by a redemption notice is not strong; it is merely slow to leave.

Governments compound the asymmetry. Their 518,500 coins are largely the product of enforcement seizures, not conviction purchases. When Germany liquidated its holdings in 2024, it was simultaneously a regulatory act and a supply shock — an exogenous, essentially unpredictable sale. A holder whose behavior is governed by fiscal need rather than price is the opposite of a price-insensitive buyer. If the United States were to formalize a strategic bitcoin reserve and forbid sales, the meaning of government holdings would invert from latent pressure into locked supply. But that remains speculation, not policy.

Put these pieces together and a reflexive loop comes into view. Corporate treasuries sell; price falls; net asset values shrink; the mNAV premium narrows; financing capacity declines; forced selling follows. Each step feeds the next. The 2022 collapse of FTX, during which I retreated to a cabin on Hiiumaa island for six weeks and wrote The Hollow Promise of Yield, taught me that the most dangerous structures are the ones everyone has agreed not to look at. The corporate treasury flywheel is, for now, in that category. It is not a fraud. It is a lever, and levers reverse.

I do not say this as a skeptic of institutions. In 2024 I stood in a closed-door panel in Geneva, presenting a deck titled Beyond Speculation: Blockchain as a Trust Layer to institutional allocators, and I negotiated with three major asset managers to adopt a Green-DAO reporting standard for their crypto holdings. I believe institutional capital can be disciplined by ethical frameworks. But discipline requires transparency, and the current ownership data offers opacity dressed as clarity.

Contrarian Angle

Here is where I part company with the report's own framing. The title stages a war: Wall Street versus Ordinary People. It invites us to pick a side and to cheer for the little guy's 66%. But the more I look at the ledger, the more I suspect the war is a distraction from the only question that has ever mattered in this industry.

Ownership is the wrong axis. Decentralization was never about the distribution of coins; it was about the distribution of the power to verify. A system can be owned by millions and verified by three. A system can be owned by three and verified by millions. The first is a democracy in name and an oligarchy in fact. The second is the opposite, and it is closer to what bitcoin actually is. On that measure, the 66% figure is not merely inflated. It is irrelevant.

This is why I have grown impatient with the celebration of ETF inflows. When bitcoin became an ETF, it gained a price feed to the world's largest capital pools — and it lost something quieter. The chain that Satoshi described as peer-to-peer electronic cash now settles most of its economic weight inside custodial wrappers whose movements never touch the base layer. The base layer's own activity — the fees miners earn, the blockspace actually consumed — has become a rounding error next to the paper claims stacked on top of it. A monetary system whose price is set by macro liquidity and whose settlement happens in custody is no longer a peer-to-peer system; it is a commodity with a blockchain attached. The ETF did not kill bitcoin's cypherpunk ambition by force. It killed it by convenience.

The same substitution of appearance for substance runs through the infrastructure. In DeFi, we still pretend that price oracles are decentralized. Many oracle networks achieve their resilience by running a small set of permissioned nodes whose operators are, in practice, as concentrated as the validators they were meant to replace. Latency — the gap between the real world and the feed — remains the sector's structural Achilles' heel, and no amount of token distribution hides it. On the rollup side, the cost of generating zero-knowledge proofs is still high enough that several operators are subsidizing verification out of venture capital rather than out of revenue. When the subsidy stops, the security assumptions get renegotiated. Verification, not ownership, is where the fragility lives.

This reframes what we should be measuring. The report asks how many coins ordinary people hold. The better question is how many people can independently verify the state of the system they depend on. On that measure, the numbers are grim. Most holders — retail and institutional alike — cannot audit the custody arrangements that hold their coins. They trust a custodian, a filing, a dashboard. They have outsourced verification to exactly the intermediaries that bitcoin was built to remove. The most centralized thing about modern bitcoin is not its mining; it is its accounting.

I carry this concern into my current work. This year I helped design a decentralized identity protocol for Tallinn's AI startup hub, integrating zero-knowledge proofs into AI agent wallets so that autonomous agents can prove their origin without revealing proprietary data. One hundred agents piloted the system, moving five million dollars in secure transactions. The lesson I took from it is the one I keep returning to in every domain: identity, ownership, and value all reduce to the same primitive — the ability to prove a claim without trusting a gatekeeper. When that primitive weakens, no distribution of coins can compensate.

Takeaway

Silence is the first vote in a true consensus. The phrase returns because the ledger has been silent on the question we most need answered: not who holds the coins, but who can prove they hold them, and under what terms they might let go.

The 66% figure is not a lie so much as a category error dressed as a fact. Subtract the lost, the ghost, and the custodians, and the free float is smaller and more fragile than the headline suggests. The institutional buyers who absorbed the difference are not saviors; they are levered, concentrated, and, on the evidence of Strategy's own cash flows, beginning to feel the strain of maintaining what they built. The transfer of coins from individuals to institutions is real. Whether it strengthens bitcoin or merely moves its fragility to a higher floor is the question this cycle will answer.

I will be watching three numbers: the mNAV premium on the largest treasury holders, the net flow into the ETFs, and the on-chain movement — or continued stillness — of Satoshi's coins. Each is a vote, and each, for now, is cast in silence. The first vote in any true consensus is always the quietest. The question for the next twelve months is whether that silence is patience or paralysis — and who, when the counting finally begins, will be able to verify the tally.