Which companies hold crypto on their balance sheet? The 2026 map

What this page is, and what it is not

This is a dated snapshot of listed companies that hold cryptocurrency as a treasury asset, compiled on 26 August 2026 from the filings named against each figure. It is not a league table, not a ranking of quality, and explicitly not a buy list. Naming a company here is not an endorsement of it, its shares, its preferred stock or the token it holds. We publish no price views and no targets.

Treat the numbers as perishable. Several of these companies disclose holdings weekly by Form 8-K, so a figure can be out of date within days of publication. Strive bought 1,110 bitcoin in a single five-day window in August 2026; BitMine says it has bought ether every week since June 2025. Any list you find without an "as at" date on every row — including this one, if you read it in six months — is telling you what was true on some unknown day.

The names also move. At least nine companies in this cohort renamed themselves between mid-2025 and mid-2026, and in most cases the ticker did not change with the name: MicroStrategy is Strategy but still MSTR; Asset Entities is Strive but still ASST; SharpLink Gaming is Sharplink but still SBET; Marathon Digital is MARA Holdings; Riot Blockchain is Riot Platforms; Helius Medical Technologies is "Solana Co" but still trades as HSDT, a ticker that reflects a neurostimulation-device past. A list more than about three months old will contain companies that no longer exist under that name and tickers that now point somewhere else.

This page is the map. The mechanism — how a treasury company is supposed to create value, and the exact point at which the machine runs backwards — is set out separately in our guide to how digital asset treasury companies actually work.

The risk framing you need before the table

A digital asset treasury company is not a wrapper that gives you the coin. It is an operating company, usually with debt and preferred stock ranking ahead of you, whose share price is the coin price multiplied by whatever premium or discount the market currently applies. When that multiple contracts, the shares fall further than the coin.

That is not a hypothetical. Bloomberg reported on 5 June 2026 that bitcoin treasury companies had shed roughly $62bn of combined market value between early October 2025 and early June 2026, with aggregate capitalisation falling from about $134bn to about $72bn — and that share-price losses in many cases exceeded the losses on the bitcoin actually held. Strategy's own investor materials put its one-year total shareholder return at approximately minus 65% as at 21 August 2026, a period in which bitcoin fell far less. By late December 2025, DL News reported that of roughly 195 treasury companies then in existence, only a handful traded above the value of their own coins.

Leverage makes this sharper. Riot has 5,821 bitcoin pledged against a $200m credit facility; Nakamoto has 3,805 of its 4,467 bitcoin pledged to a Kraken loan; BTCS borrows USDT against its ether on the AAVE protocol, where liquidation is automatic and there is no lender to negotiate with. Galaxy Research warned during 2026 that at least five digital asset treasury companies could be forced to sell assets, merge or shut down that year — that is an analyst view, reported secondhand, not a fact, but it describes a mechanism the filings support. The full anatomy is set out in what actually goes wrong at a bitcoin treasury company.

On protection: none of this is covered by the Financial Services Compensation Scheme. The FSCS does not compensate anyone for investment losses, and most of the securities below are listed in the United States or Japan and sit entirely outside UK compensation arrangements for market risk in any case. If the coin falls, or the premium collapses, or the company restructures, the loss is yours. See the FSCS for what it does and does not cover.

Who holds what, as at August 2026

Read the "holdings" column with the date and the source, never on its own. We grade each figure: filing means a number we took from a company's own SEC filing; company release means the company said it but we could not tie it to a filing; tracker means a third-party aggregator; contested means published figures disagree. Dollar figures come from US filings and are left in dollars; sterling figures come from UK sources. We have not converted anything.

Scale, and the concentration you must not miss. Bitwise counted 174 public companies holding a combined 1,187,898 BTC as at 12 May 2026 — but Strategy alone accounted for 818,869 of them, roughly 69% of the entire corporate pool. Any statement about "corporate bitcoin holdings" is overwhelmingly a statement about one company. A live read of bitcointreasuries.net accessed on 25–26 August 2026 showed 365 tracked entities holding $346.68bn of digital assets, of which 215 public companies held $121bn; that page displays no explicit as-of date, and The Block's corporate tracker showed only 119 companies as at 28 July 2026. The trackers disagree by a factor of two because they count different things.

European names we are deliberately not putting numbers on. Bitcoin Group SE (Germany), Capital B (France, formerly The Blockchain Group), H100 Group (Sweden), Treasury B.V. (Netherlands) and Sequans Communications (France) all appear on sector lists with holdings figures. We could not date or verify any of those figures against a primary regulatory announcement in the time available, so we are not publishing them. That is the honest answer, and it is also the point of this page.

How these companies pay for the coins

The funding method determines how the company behaves in a fall, so it matters more than the headline holding. There are four distinct forms, in roughly ascending order of danger.

1. At-the-market equity. Continuous issuance of ordinary shares straight into the market. Strategy sold $2,006.5m of stock in the week to 23 August 2026; Strive issued 3,646,300 Class A shares in the week to 21 August. This is only accretive per share while the stock trades above the value of the coins. It carries no forced-sale trigger, but it dilutes relentlessly, and the share count does not shrink when the premium disappears.
2. Perpetual preferred stock. Strategy's STRK, STRF, STRD, STRC and STRE; Strive's SATA; BitMine's 9.50% BMNP; and the sterling preferred the Smarter Web Company has been preparing since its July 2026 capital reduction. These shift risk onto a new class of income-seeking buyers. The dividend is not free: it must be paid in cash, and where the operating business cannot generate it, the cash comes from issuing more shares or selling coins.
3. Convertible and secured debt. Riot's $200m Coinbase Credit facility secured on 5,821 BTC; Nakamoto's Kraken loan secured on 3,805 of its 4,467 BTC at 7.75%–8.00%; Satsuma's £163.6m of convertible notes, half its treasury sold to repay them; DFDV's $135.3m of digital-asset financing borrowings in six months. Debt turns a price fall into a collateral call.
4. On-chain DeFi leverage. BTCS borrowing $53.0m of USDT against 46,531 ETH on AAVE. This is the purest form: a falling price mechanically liquidates the treasury, at protocol speed, with no lender on the other end of a phone.

Categories three and four are the ones that convert a bad quarter into forced selling. Category two is the one most likely to be mis-sold to a retail income buyer, because a double-digit perpetual dividend on a company whose only real asset is a volatile commodity is not a bond and should not be read as one.

What does "coins held" actually mean?

This is the most under-reported risk in the sector. A headline holdings figure is not a claim on unencumbered assets. From the 30 June 2026 filings:

• MARA: 9,270 of 35,577 BTC — 26% — loaned or pledged.
• Riot: 5,821 of 11,380 BTC — 51% — pledged to a credit facility.
• Nakamoto: 3,805 of 4,467 BTC — 85% — pledged to a Kraken loan.
• CleanSpark: a further 1,719 BTC recognised as a receivable from derivative counterparties, not as bitcoin at all.
• DeFi Development Corp: the majority of its Solana sits in "pledged as collateral" or "locked" buckets rather than freely available, and a further 621.9 thousand SOL delegated to its validators by customers is not company property at all.

There is a second trap in the accounting. Directly held crypto is now fair-valued through income under US GAAP, so losses reverse if prices recover. But liquid staking tokens can be treated as intangibles subject to impairment that is never reversed. Sharplink says so in its own results: a $76.1m impairment on its LsETH and weETH positions "reduce[s] the carrying value ... under U.S. GAAP and [is] not reversed for subsequent market recoveries", while noting the charges do not reduce the number of tokens held. Reported book equity for a company holding staked or wrapped tokens is therefore a ratchet that only turns one way.

What is ASST, and who owns it now?

Reporting on this ticker is unusually confused, so here it is in order. ASST is Strive, Inc. It is not Asset Entities.

• The shell. Asset Entities Inc. was a small Nasdaq-listed social-media and Discord marketing company. It was the listed vehicle, not the business.
• The reverse merger. Strive Enterprises, Inc. — the operating asset manager associated with Vivek Ramaswamy's "anti-ESG" firm Strive Asset Management — merged into that shell, completing on 12 September 2025 with approximately $750m of equity financing raised at closing and up to a further $750m available on warrant exercise. This was a reverse merger into a nano-cap listed shell, not an IPO.
• The registrant today. SEC CIK 1920406 now reads Strive, Inc., with "Asset Entities Inc." recorded as a former name until 10 September 2025. It is Nevada-incorporated, based at 200 Crescent Court, Dallas, Texas, and files under Commission File 001-41612. Registered under Section 12(b): Class A common stock trading as ASST on Nasdaq, and Variable Rate Series A Perpetual Preferred Stock trading as SATA on Nasdaq (Form 8-K, 24 August 2026).
• The reverse split. A 1-for-20 reverse stock split of the Class A and Class B common stock completed on 6 February 2026, announced alongside the Semler transaction. On 13 January 2026, when Semler shareholders approved the deal, ASST fell a reported 12–15%, with coverage attributing the fall to the unexpected split rather than the acquisition.
• Semler Scientific no longer exists as a standalone treasury company. This is the most commonly garbled fact in coverage of this sector. Strive acquired Semler; the merger closed on 16 January 2026; each Semler share converted into 21.05 Strive Class A shares; SMLR was delisted from Nasdaq and deregistered. EDGAR CIK 1554859 now shows no tickers and no exchanges. Semler contributed 5,048.1 BTC, and combined holdings at closing were reported as 12,797.9 BTC. Semler is the acquiree, not a continuing peer — any 2026 list showing SMLR as a live treasury company is wrong.

Holdings. Strive reported 21,356 bitcoin as at 21 August 2026, up from 20,246 BTC at 14 August, having bought 1,110 BTC between 17 and 21 August at an average of approximately $73,409 per coin inclusive of fees. Cash and equivalents were $171.9m. Source: the Form 8-K of 24 August 2026 linked above.

The funding engine, in one week. In that same single week Strive issued 3,646,300 new Class A shares and 441,313 new SATA preferred units. Class A shares outstanding were 79,890,888, Class B 9,792,535, SATA 8,270,815, assumed fully diluted 92,949,226, with a further 26,596,010 shares underlying traditional warrants. That is the at-the-market machine at full tilt: sell paper above the value of the coins, buy coins, repeat, weekly.

Why do the miners hold less bitcoin than they used to?

Because several of them have stopped accumulating, and some are net sellers. This is the quietest important shift of 2026 and it is plainly visible in the filings.

MARA's holdings fell 29% year on year, to 35,577 BTC at 30 June 2026, with 9,270 of those coins loaned out or pledged. The company mined 2,422 BTC in the quarter and posted adjusted EBITDA of negative $360.9m, against positive $1.2bn in the same quarter of 2025. Its own language has moved: it now describes itself as "a leading digital infrastructure company" deploying "digital energy technologies".

Riot's unrestricted stack collapsed. The balance-sheet line for bitcoin fell from $1,227.5m at 31 December 2025 to $325.4m at 30 June 2026, with restricted bitcoin roughly flat at $340.7m — that is, more than half of what it still holds is pledged. Meanwhile it contracted 241 MW of AI data centre capacity with AMD and an unnamed frontier AI lab: a 191 MW lease at Rockdale with a 20-year initial term to June 2048, approximately $9.1bn of initial contract value and estimated average annual net operating income of $365–411m, financed initially by a $573m interim facility from Morgan Stanley. Its cash cost to mine a bitcoin in Q2 2026 was $49,912 excluding depreciation. The point for a reader: the bitcoin balance sheet has become the collateral and seed capital for a completely different business.

CleanSpark is underwater and buying back its own stock. It held 12,205 BTC at a cost basis of $90,188 per coin against a $58,524 fair value at 30 June 2026, reported a $239.8m quarterly net loss, watched its accumulated deficit rise to $1.153bn from $125.9m, and lifted treasury stock to 42,365,391 shares ($608.161m) from 11,759,935. It now calls itself "a market leading data center developer" and has signed a 20-year $6.6bn triple-net lease. Its own chief financial officer opened with "Despite currently challenging bitcoin mining economics".

The buyback is not a contradiction; it is the rule working. When a company's shares trade below the value of its coins per share, issuing shares to buy coins destroys value per share and buying back shares creates it. That single ratio governs the whole sector and is worth understanding properly before you read another treasury press release — see mNAV explained.

One more distinction the tables never make: coins that are loaned or pledged are not the same asset as coins held outright. Pledged coins are already promised to somebody else in the event you most care about. Loaned coins carry counterparty risk that has nothing to do with the bitcoin price. Treating the two as interchangeable is how a "holdings" number flatters a balance sheet.

The small-cap shell pattern

The standard structure in this sector is not a new company. It is a reverse merger into an existing small-cap listed company with a failing or marginal operating business — and you can verify the whole chain yourself from EDGAR's "former names" field. A sample of the lineages, all from the SEC's own records:

• AspenBio Pharma → Venaxis → Bioptix → Riot Blockchain → Riot Platforms
• MTech Acquisition (a SPAC) → Akerna Corp (cannabis compliance software) → Gryphon Digital Mining → American Bitcoin
• KBL Merger Corp IV (a SPAC) → 180 Life Sciences (biotech) → ETHZilla → Forum Markets
• Volcon (electric powersports) → Empery Digital
• SRM Entertainment (toy manufacturer) → Tron Inc.
• Surna Inc. (cannabis cultivation equipment) → CEA Industries
• Helius Medical Technologies (neurostimulation devices) → "Solana Co", still trading under HSDT
• Eyenovia (ophthalmic pharma) → Hyperion DeFi, still filing under a pharmaceutical industry code
• Nth Games → Super League Gaming → Super League Enterprise, the shell Metaplanet agreed in August 2026 to seed with 2,100 BTC

The pattern is the point. The listing is the asset being recycled; the business inside it is whatever narrative is raising money that year. ETHZilla is the cleanest single example: one legal entity, four identities in six years — SPAC, then biotech, then ether treasury, then tokenisation platform — each matched to the story of the moment.

What typically happened to the share price. A violent re-rating on announcement, then collapse once the premium to net asset value closed, frequently ending in a minimum bid price deficiency notice and a reverse stock split. Documented in this cohort: Nakamoto fell 99% from peak, received a Nasdaq delisting notice in December 2025 and executed a 1-for-40 reverse split on 22 May 2026, cutting shares outstanding from approximately 696.1m to approximately 17.4m. Upexi received a bid price notice on 30 July 2026. Strive did a 1-for-20 reverse split in February 2026. Satsuma fell more than 99% from its June 2025 peak before being voted into liquidation. Semler was reported down 74% during 2025 before being absorbed. The Smarter Web Company fell 22.5% in early trading in its first days as a Main Market stock.

We are stating that qualitatively, from named cases, on purpose. We found no rigorous study of average abnormal returns around treasury-adoption announcements, so any "on average these shares did X%" claim you see should be treated as invented until it names a method.

How should you treat a treasury announcement?

As information about a company's financing plans, not as information about the coin. A useful set of questions, all answerable from public documents before the excitement wears off:

• What was this company last week? Pull the EDGAR former-names record. If the operating business is a failing esports, cannabis, toy or medical-device company, the listing is the product.
• Where is the money coming from? Equity, preferred, secured debt or on-chain leverage. Only the first has no forced-sale trigger.
• What is the share count doing? Compare the cover page of the latest filing with the one before it. Nakamoto's pre-split 696m shares are the whole story of what accumulation cost its original holders.
• Is the company issuing above or below the value of its coins? Sharplink states plainly that it issues only at a premium to net asset value and otherwise buys back stock. Companies that keep issuing below it are transferring value away from existing holders.
• How much of the treasury is actually free? Look for "pledged", "restricted", "loaned", "locked" and "collateral" in the digital assets note.
• What must be paid in cash, whatever happens? Preferred dividends and interest do not pause when the coin falls.
• Does the headline metric survive contact with the definition? See the next section.

The failures that lists leave out

Almost nobody publishes this section, which is why it is the most useful one. These are companies that wound down, abandoned the strategy, were acquired, or lost their listing. Each is dated and each names what actually happened.

Read that table against the marketing language of 2025, when bitcoin on a corporate balance sheet was routinely described as a permanent reserve asset that would never be sold. Within about a year, Strategy was selling below cost to fund preferred dividends, Empery had sold half its stack, Bitdeer had reportedly sold all of it, Satsuma had voted to sell everything and return the cash, and Semler had ceased to exist. The "never sell" framing did not survive contact with a drawdown.

How to verify any of this yourself

A snapshot goes stale. A method does not. Everything on this page came from documents you can open, and here is the order to open them in.

1. Find the registrant, not the brand. Search the company on EDGAR full-text search or the EDGAR company browser and note its CIK number. The CIK never changes, even when the name and ticker do.
2. Read the former names. The SEC's submissions record for a CIK carries a formerNames field — for example https://data.sec.gov/submissions/CIK0001920406.json for Strive. This is the fastest honest history of what a company used to be, and it is where the shell chains in this page came from.
3. For the current holding, read the weekly Form 8-K. Strategy, Strive and BitMine all disclose under Item 8.01. The holding, the average price, the share counts and the cash balances are usually in one table.
4. For the real balance sheet, read the 10-Q digital assets note. This is where you find cost basis versus fair value, and — crucially — the split between unrestricted, restricted, pledged, loaned and locked. Search the document for "pledged", "collateral", "restricted" and "locked".
5. Compare the press release with the balance sheet. They frequently count different things. DeFi Development Corp's headline 2,294,576 SOL and its 888 thousand fair-valued SOL units are both true statements about different buckets.
6. Check the share count on the cover page. Holdings up and shares up faster means fewer coins per share. This is the number a "coins per share" metric is supposed to capture and often does not.
7. Know who does not file quarterly. Foreign private issuers such as Bitdeer and Sequans file Form 6-K and are not required to file 10-Qs, so their disclosure cadence is slower and cross-company comparison is weaker than it looks.
8. For UK companies, use the regulatory feed, not the press. Company announcements are filed as RNS and held on the FCA's National Storage Mechanism. For Japanese issuers such as Metaplanet, the equivalent is TDnet.
9. Check corporate actions. Reverse splits, ticker changes and delistings are published by the exchanges. This is how you catch that a ticker in an old list now points at something else.
10. Treat aggregators as a starting point only. On 25–26 August 2026 one tracker showed 365 entities and $346.68bn with no visible as-of date, while another showed 119 companies as at 28 July 2026. Neither is wrong; they count different universes. Always click through to the filing.

What this means if you are a UK investor

Three practical points, none of which are recommendations.

These are shares, taxed as shares. Buying MSTR, ASST or BMNR is buying a foreign equity, not a cryptoasset. The disposal is a share disposal for Capital Gains Tax purposes and the pooling rules for shares apply, not the cryptoasset guidance you would use for holding bitcoin directly. Dividends on the preferred instruments are foreign dividends. Strategy expects its own preferred distributions to be treated as a non-taxable return of capital for US federal tax purposes because it expects no earnings and profits — that says nothing whatever about UK treatment, which follows separate rules.

The wrapper question is separate from the exposure question. If what you actually want is exposure to a coin rather than to a leveraged listed company sitting on top of one, the rules on crypto exchange-traded notes, and whether they can go inside an ISA, are covered in crypto ETNs and the UK ISA rules. The FCA's own material on cryptoassets is at fca.org.uk.

No compensation scheme covers the outcome. The FSCS does not protect you against investment losses. It is not a backstop for a coin falling, a premium collapsing, a preferred dividend being suspended, or a company restructuring its obligations. Where these securities are listed overseas, UK protections are further removed still. Anyone implying otherwise is wrong.

Sources and the limits of this snapshot

Every holding figure above is attributed to a document and a date. Where we could only find a company press release, a third-party tracker, or conflicting reports, we have said so in the row rather than smoothing it into a clean number. Where a figure could not be verified at all — Upexi's current Solana position, the European treasuries, the Twenty One Capital total — we have left the gap visible.

Market context for the dates used, so you can see how far the marks have moved: bitcoin was fair-valued at $58,524 at 30 June 2026 in both MARA's and CleanSpark's accounts; Strategy sold at an average $64,262 in the week to 9 August 2026; Strive bought at an average $73,409 between 17 and 21 August 2026; and market reporting put bitcoin's opening price at $78,982.27 on 25 August 2026 with an intraday high of $81,023.41, and ether's opening price at $2,482.37. Those are dollar figures from US filings and US market reports and we have not converted them.

What this page deliberately does not do: forecast any price, rank these companies, describe any of them as an opportunity, or suggest an allocation. It is a description of what a set of public documents said on a particular day about a fast-moving, speculative and heavily promoted corner of the market. If you are considering any of this with money that matters, take regulated advice.

Related guides on digital assets and tax

The four companion pages on treasury companies are linked in context above: the mechanism explainer, mNAV, what goes wrong, and crypto ETNs and ISAs. Elsewhere on the site:

• What are digital assets? A UK primer
• Crypto CGT in the UK
• Crypto tax calculator
• Complete UK CGT guide
• UK ETF guide
• Investment scam checklist
• Investing hub