Volume III · Bitcoin · Money Research

03 — How Bitcoin Is Actually Used: Global Adoption, 2009–2026

The headline numbers

Bitcoin appears in individual wallets, brokerage funds, company treasuries, payments and public-sector asset decisions. These channels do not yield one verified worldwide owner count: a cryptocurrency-user estimate is not a Bitcoin-owner estimate, and an ETF shareholder, trust asset and custodian key are overlapping layers. A comparable government-owned total also cannot be established by adding labelled wallets: seizure, final forfeiture, state-company ownership, fund investment and private declarations have different legal statuses. The current global-headcount, ETF/corporate-total and market-share figures formerly used here are withheld pending same-date, non-overlapping definitions and exact source observations. The question is not simply whether bitcoin has been adopted but as what.

Survey use and Bitcoin-specific limits

The strongest directly reviewed household measure is narrower than a worldwide Bitcoin-use estimate. In the Federal Reserve's 2025 survey, 10% of U.S. adults reported any cryptocurrency use, 9% reported buying or holding cryptocurrency as an investment, 2% reported buying something or making a payment with it, and 1% reported sending it to friends or family. Categories could overlap. These percentages do not give the fraction of Bitcoin owners who spend Bitcoin. Federal Reserve, 2025 SHED, Banking, table 33. Exchange flows, on-chain transfer counts, mining fees and payment use measure different activities; the figures cited elsewhere in this chapter should not be combined into a single global payment share without compatible populations and methods.

Exchange balances, dormant-output counts and fund/corporate custody labels are not one measure of beneficial holders or liquid float. The earlier 2023–26 exchange and inactivity figures are withheld here because their trackers, address/entity rules and snapshot times are not a reconciled primary series. A coin left unmoved for five months remains an unspent output; it is not necessarily lost, held for investment or outside possible trade. Bitcoin Core, `gettxoutsetinfo` scope.

Who adopts it, and where

A crypto-activity index ranks activity under its own asset, channel and purchasing-power method; it is not a map of Bitcoin owners or payment users and cannot identify every participant's motive. The earlier country rank and regional-dollar totals from industry indexes are withheld here pending exact report versions, covered flows, observation windows and asset splits. Nigeria below supplies a narrower example with two explicitly different denominators. Chainalysis, 2025 Sub-Saharan Africa report, methodology and Nigeria section.

Nigeria shows why the activity measured matters. Chainalysis reports that Bitcoin accounted for 89% of fiat purchases of crypto on the centralized exchanges it covered in Nigeria, while USDT accounted for 7%. Its 2025 geography report expressly excludes informal, business-to-business and over-the-counter transfers from that purchase comparison; the 89% is not a share of Nigerians, payments or all crypto flows. Chainalysis, “Bitcoin holds dominance”. Separately, the IMF's 2026 Nigeria report estimates that USDT and USDC made up over 65% of Nigeria's crypto inflows in 2024, using an on-chain cross-border-flow proxy available through January 2025. That is a different denominator and transaction channel, not a contradiction or proof that one asset dominated every use. IMF Country Report 26/125, annex VII, pp. 64–65.

Bitcoin can be held as an investment, used for transfers or accessed through a regulated product, but those channels do not identify a universal motive by national income group. U.S. retirement-plan exposure depends on fiduciary choices: Executive Order 14330 directed Labor guidance review and SEC consideration of specified alternative-asset access, and Labor's March 2026 item was a proposed rule, not a mandate that every plan offer Bitcoin. Claims about sanctioned users, remitters or stablecoin preference need transaction and population evidence rather than inference from geography. Executive Order 14330, §§2–3; Labor proposed rule, Federal Register, 31 March 2026.

Institutional and sovereign adoption

U.S. spot Bitcoin exchange-traded products and FASB fair-value rules changed routes to exposure and accounting for qualifying corporate holdings; they do not by themselves explain every company's decision. The former 31 August 2026 Strategy balance, listed-company aggregate, named fund positions and sovereign-allocation quotations are withheld pending exact SEC/manager filings, non-overlapping ownership definitions and statement transcripts. The Czech National Bank announced a mixed-digital-asset test portfolio outside official reserves, which is not a central-bank reserve allocation or a worldwide first-purchase census. CNB test-portfolio release; FASB ASU 2023-08, scope and effective date.

The government record is heterogeneous. The US reserve order covers eligible finally forfeited Treasury bitcoin, not every seizure or forfeiture claim; the Prince Group announcement was a complaint, not a final order. UK prosecutors report recovery of about 61,000 bitcoin still subject to civil proceedings. China's PlusToken disposition is not verified by the cited wallet trackers, and Bhutan's documented state-linked mining does not make attributed wallet outflows proven sales. El Salvador's purchases, a sovereign fund's ETF investment and a central bank's test portfolio are also different forms of exposure. “Countries That Adopted Bitcoin, in What Form, and What Happened” separates the cases and their legal status. US reserve order; DOJ Prince Group announcement; UK CPS.

The industry behind it

Mining turns energy, hardware and operating capital into block rewards, which include the programmed subsidy and transaction fees. Hashrate is an estimate from observed block timing and difficulty, not a direct census of miners or countries. Hashrate Index's 14 September 2026 weekly report estimated a 943 EH/s seven-day average and a 928 EH/s thirty-day average; it reported a 0.0183 BTC seven-day average of fees per block, alongside the 3.125 BTC subsidy. These are dated network estimates, not a miner's net dollar profit or an attack-cost calculation. Energy use, country shares, survey coverage, AI-hosting contracts and claims of machine shutdown require separate definitions and exact observations before being promoted as a single industry trend. Hashrate Index, 14 September 2026 weekly roundup, “Hashrate & Difficulty” and “Transaction Fees”.

Exchanges, custodians, market makers, stablecoin issuers and fund sponsors form distinct service layers around mining and ownership. The former Coinbase/Tether quarterly revenue, Treasury-asset, coin-balance, ETF-custody and sponsor-fee totals are withheld here until the named entity's original filing or assurance report, dates, units and consolidation scope are accepted. Assets held for a fund or client cannot be counted as the custodian's own beneficial coins. iShares Bitcoin Trust ETF, June 2026 Form 10-Q.

Payments: the small, real corner

Bitcoin payments exist through merchants, remittance providers and Lightning wallets, but a global Bitcoin-specific merchant-spending share has not been established here. Lightning's public channel capacity measures funds advertised in public channels, not completed payment value, private-channel balances, route success or the asset paid; a custodial withdrawal count measures a provider's own users. The former 12-million-transaction, Coinbase-withdrawal, $380-million-capacity and global-payment-value figures need dated operator reports with compatible periods and definitions. Gross stablecoin transfers likewise include trading, exchange and internal movements rather than only purchases. A comparison with card or remittance rails must match the payer-to-recipient journey, fees, conversion, failure and recourse. Lightning Labs, “Managing Channel Liquidity”; Visa Onchain Analytics, adjusted transaction methodology.

How bitcoin has behaved when it mattered

The crisis record cannot be compressed into one return or causal story. For a precisely dated example, the Grayscale Bitcoin Mini Trust ETF's June 2026 filing valued one bitcoin at $87,549.41 on 31 December 2025 and $58,745.18 on 30 June 2026, using its principal-market price at 4 p.m. New York time. That demonstrates a substantial decline over this six-month window; it is not a price path for every venue, an investor's after-fee return, or proof that a particular war caused the move. Grayscale Bitcoin Mini Trust ETF, Form 10-Q, note 3. Gold also fell sharply from a January 2026 intraday high to late June during the Middle East shock, according to the World Gold Council, but those two cited price conventions are not a matched return series. WGC, *Gold Mid-Year Outlook 2026*, “A rollercoaster ride”. Earlier 2020–25 anecdotes, the late-August correlations and year-on-year rankings are withheld until the same source, dates, currency and cost treatment can be reproduced across assets.

These selected episodes show large bitcoin drawdowns and rebounds, but they are not a matched test of inflation protection or crisis performance. The timing of a shock, the investor's currency, the comparison asset and the holding window can change the result. An IMF staff study found stronger Bitcoin–equity co-movement after 2020 and a broader crypto factor sensitive to modeled U.S. monetary tightening in its 2018–23 sample. That is evidence against assuming an automatic market-risk hedge, not proof that bitcoin behaves identically in every later crisis. IMF Working Paper 23/163, introduction, table 3 and conclusion. A reliable “debasement” or safe-haven claim needs dated, comparable return and drawdown series rather than a narrative selection of episodes.

Key takeaways

Investment and trading are prominent uses of Bitcoin, but a defensible global split between holding, trading, payments and transfers has not yet been established here. The 2025 Federal Reserve survey provides U.S.-adult percentages for cryptocurrency, not Bitcoin-specific user shares: 9% reported investment use, 2% purchase/payment use and 1% sending to friends or family, with overlapping answers. Nigeria illustrates the measurement problem: Bitcoin's 89% share of covered centralized-exchange fiat purchases and dollar stablecoins' over-65% share of estimated 2024 crypto inflows describe different channels. Institutional holdings, regulated access and custodial products add other forms of exposure that should not be counted as unique owners without adjustment. Differences across countries and crises need their own dated, comparable evidence before this chapter can claim a universal adoption pattern or a reliable hedge function. Federal Reserve, 2025 SHED, Banking, table 33; Chainalysis, 2025 Nigeria purchase comparison; IMF, 2026 Nigeria report, annex VII.