“27% of all the Bitcoin in circulation was controlled by roughly 0.01% of holders which is about 10,000 entities.”

• The claim closely tracks a specific, widely reported academic finding: Igor Makarov (LSE) and Antoinette Schoar (MIT Sloan), 'Blockchain Analysis of the Bitcoin Market,' NBER Working Paper 29396 (Oct. 2021), which estimates that the top 1,000 individual investor entities controlled roughly 3 million BTC and the top 10,000 roughly 5 million BTC.
• 5 million BTC measured against the ~18.9-19 million BTC in circulation at the end of 2021 is ~26-27%, which is exactly the '27%' figure in the claim — the arithmetic is internally consistent with the source study.
• The Wall Street Journal's December 2021 write-up of the paper framed the result in almost the same words the claim uses: roughly 0.01% of bitcoin holders (about 10,000 entities) control about 27% of circulating bitcoin, and the top 10,000 investors hold roughly a third of coins held by identifiable individuals.
• Independent on-chain data is directionally consistent with extreme concentration: public rich lists show the top ~100 bitcoin addresses alone hold on the order of 13-15% of all coins, so 10,000 entities holding a quarter-plus is plausible rather than an outlier estimate.
• The authors used address-clustering (heuristics that group addresses into single controlling entities) and explicitly separated intermediaries such as exchanges and miners from individual investors, so the '10,000 entities' figure is not simply a naive count of top addresses.

• The figure is a snapshot of late-2021 blockchain data, not a current fact. Circulating supply is now ~19.9 million BTC and custody has shifted substantially (spot bitcoin ETFs launched in 2024 now hold well over a million BTC through a handful of custodial entities), so the exact 27% number should not be quoted as present-day without a date qualifier.
• The '0.01% of holders' denominator is the weakest part of the statement. It implies about 100 million holders, a rough industry estimate; other estimates of crypto owners range from ~100 million to well over 400 million, which would make the share 0.002-0.01%. The percentage is an approximation layered on an approximation.
• 'Entities' are clusters of addresses inferred by heuristics, not verified people. One entity can be a custodian, a mining pool, an exchange cold-wallet cluster, a lost-key hoard (e.g., Satoshi-era coins) or an ETF trust representing millions of beneficial owners; conversely one wealthy person can be spread across many clusters. Clustering can both overstate and understate individual concentration.
• The headline framing can mislead in the opposite direction too: because exchanges and other intermediaries were treated separately, the coins sitting in custodial wallets on behalf of retail users are not all inside that 27%. Depending on how you count intermediaries, concentration of *control* can be even higher than 27%, or concentration of *beneficial ownership* somewhat lower.
• The primary source is a working paper (subject to revision, and versions report slightly different cuts, e.g., 'about one-third of bitcoin held by individuals' vs. '27% of all bitcoin in circulation'), so precise percentages vary by which version and which baseline is cited.

84/100 — the lab’s own confidence in the score above, not a second reading of the claim.

National Bureau of Economic Research (Makarov & Schoar, 'Blockchain Analysis of the Bitcoin Market')journal · 92 · NBER (full working paper PDF)journal · 92 · The Wall Street Journalnews · 88 · BitInfoCharts (on-chain address rich list)other · 62 · Blockchain.com (circulating supply data)other · 70

• The claim closely tracks a widely reported finding from Igor Makarov and Antoinette Schoar’s NBER/Management Science research on Bitcoin ownership concentration: after clustering addresses into likely entities and separating intermediaries such as exchanges, they estimated that the top 10,000 individual Bitcoin investors controlled about 5 million BTC. Against roughly 18.5–19 million BTC then in circulation, that is about 27%.
• The Guardian reported the same result in near-identical language: the top 0.01% of Bitcoin holders controlled 27% of the currency, citing the Makarov/Schoar study.
• The Wall Street Journal likewise reported that Bitcoin ownership was highly concentrated and that the top 10,000 accounts/entities held a very large share of outstanding Bitcoin, citing the same academic work.

• The wording 'holders' is imprecise. Bitcoin blockchain addresses are not the same as people or beneficial owners; one person can control many addresses, and custodians/exchanges can control addresses on behalf of many customers. The Makarov/Schoar figure is an estimate based on clustering and classification, not a direct census of all holders.
• The statement omits timing. The underlying estimate was based on historical blockchain data around the end of 2020/early 2021, so it should not be read as a necessarily current figure.
• The phrase 'controlled' can be misleading if interpreted as beneficial ownership rather than ability to move coins from clustered wallets. Custodial arrangements, lost coins, and dormant wallets complicate the meaning of control.

84/100 — the lab’s own confidence in the score above, not a second reading of the claim.

National Bureau of Economic Researchother · 91 · Management Science / INFORMSjournal · 92 · The Wall Street Journalnews · 88 · Communications of the ACMjournal · 86

• The claim accurately reflects the findings of a widely cited 2021 working paper (No. 29396) published by the National Bureau of Economic Research (NBER), authored by researchers Antoinette Schoar (MIT) and Igor Makarov (London School of Economics).
• The researchers utilized network clustering algorithms to map millions of anonymous Bitcoin addresses to distinct 'entities', carefully attempting to separate intermediary entities (like large cryptocurrency exchanges that hold funds for millions of retail users) from individual investors.
• In December 2021, The Wall Street Journal directly reported on this study, explicitly stating the statistic: 'roughly 0.01% of bitcoin holders control 27% of the 19 million bitcoin in circulation,' noting this concentration represented about 10,000 individual entities holding roughly 5 million bitcoins.

• Blockchain intelligence firms (such as Glassnode) frequently argue that wealth concentration metrics in crypto are difficult to measure perfectly because heuristic clustering algorithms have margins of error. It is possible that some large entities classified as 'individuals' are actually unidentified corporate custodians, trading desks, or institutional funds.
• The 27% figure does not adjust for 'lost' Bitcoin. It is estimated that 3 to 4 million Bitcoins are permanently lost due to forgotten private keys. If a large portion of these dormant whale wallets from Bitcoin's early days are included in the 'top 10,000' entities, it exaggerates the active concentration of wealth.
• The claim represents a snapshot in time from late 2021. Cryptocurrency supply distribution fluctuates, and as adoption continues to scale, retail holdings tend to slowly dilute the percentage held by early 'whale' adopters.

95/100 — the lab’s own confidence in the score above, not a second reading of the claim.

The Wall Street Journalnews · 85 · National Bureau of Economic Research (NBER)journal · 90

• The claim traces to Makarov and Schoar's 2021 NBER working paper, which estimated that the largest 10,000 individual investors controlled about 5 million BTC, approximately 27% of the roughly 19 million BTC then in circulation. Contemporary coverage by Fortune reported the same figures and attributed the research to the NBER paper by researchers at MIT and the London School of Economics.
• The arithmetic is internally consistent for the period described: 5 million divided by 19 million is about 26.3%, conventionally rounded to 27%. Fortune also reported an estimate of approximately 114 million Bitcoin accounts or users, making 10,000 approximately 0.01% of that estimated population.

• The statement is misleading as written because the underlying research estimated ownership by clustering blockchain addresses into entities; it did not directly observe 10,000 identifiable human or legal holders. Exchanges, custodians, companies, funds, and other institutions can control many addresses, while one person or entity can use many addresses.
• The 0.01% denominator was produced by combining the study's estimate of roughly 10,000 large individual investors with a separate estimate of about 114 million Bitcoin users or accounts. Those are not necessarily comparable units: the numerator concerns inferred entities or investor clusters, while the denominator is an estimated user/account population.
• The 27% figure was a historical estimate based on approximately 19 million circulating coins in late 2021, not a timeless measurement of current Bitcoin distribution. Lost coins, exchange custody, institutional custody, and unidentified ownership complicate the meaning of 'in circulation' and 'controlled.'
• The NBER paper's concentration estimates depend on address-clustering assumptions and incomplete observability. On-chain balances can show which addresses hold coins, but generally cannot prove the beneficial owner behind an address or distinguish custodial aggregation from economic ownership.
• Later address-distribution data continue to show substantial concentration, but they do not independently validate the exact claim about 0.01% of holders controlling 27%; address counts are not holder counts and should not be treated as equivalent.

89/100 — the lab’s own confidence in the score above, not a second reading of the claim.

National Bureau of Economic Researchjournal · 97 · National Bureau of Economic Researchjournal · 98 · Fortunenews · 85 · Cointelegraphnews · 72 · BitInfoChartsother · 68

• A 2021 NBER blockchain-clustering study (Makarov and Schoar) found Bitcoin highly concentrated among a relatively small set of large entities after grouping addresses; summaries of that work put control of several million BTC (on the order of a quarter of then-circulating supply) in roughly the top 10,000 non-intermediary entities, which is numerically close to the 27%/10,000 figures.
• On-chain rank data have long shown extreme inequality: a few thousand largest addresses or clustered entities hold a large minority of coins, consistent with a ~27% share at the 10,000-entity scale depending on snapshot and clustering rules.
• Global user estimates around 2020–2021 (on the order of 100 million crypto/Bitcoin-adjacent accounts) make 10,000 entities land near 0.01%, so the percentage is a plausible back-of-envelope conversion rather than an invented order of magnitude.

• The 0.01% figure requires assuming a specific total-holder denominator (~100 million); published estimates of unique Bitcoin users/entities have ranged widely (tens of millions to well over 100 million), so that percentage is not a measured quantity.
• ‘Entities’ that control keys include exchanges, custodians, ETFs, miners, and treasuries holding coins for millions of beneficial owners; treating them as 10,000 ‘holders’ overstates individual wealth concentration.
• A commonly cited ~20% of BTC is believed lost or inaccessible; those coins still appear in concentration statistics but are not controlled by anyone.
• Address-to-entity clustering is imperfect (unlinked wallets, mixing, CoinJoin), and the claim gives no date; circulating supply, ETF holdings, and exchange cold wallets have changed since the 2020–2021 snapshots that produced similar numbers.
• Public richest-address tables (which the 10,000 figure is sometimes confused with) mix exchange omnibus wallets with true whales and therefore cannot be read as 10,000 independent holders.

62/100 — the lab’s own confidence in the score above, not a second reading of the claim.

National Bureau of Economic Researchjournal · 90 · BitInfoChartsother · 72