A bitcoin treasury company is two businesses wearing one ticker: a pile of bitcoin, and a machine for issuing securities to buy more of it. The quarterly report tells you about the pile. The weekly 8-K tells you about the machine — and the machine is what changes. Strategy files most Mondays. Metaplanet files to the Tokyo Stock Exchange on a similar cadence. Smaller treasuries file when they act. If you own one of these companies, or you are trying to work out whether their buying and selling is moving the bitcoin price, the weekly filing is the primary document and everything else is commentary.
This guide is the companion to our guide on reading a bitcoin treasury company’s quarterly earnings, which covers fair-value accounting, impairment history and the income statement. This one covers the weekly disclosure: what is in it, what order to read it in, and what it means when the numbers move. Ten minutes, five numbers, two checks, four traps.
First: what a weekly 8-K actually is
Form 8-K is the SEC’s current-report form — the filing a US-listed company uses to disclose a material event between quarterly reports. Bitcoin treasury companies use it in a way most issuers do not: as a near-weekly operating bulletin covering securities issued, coins bought or sold, and dividends declared. It is short, usually two to four pages plus an exhibit, and it is free. Read the filing itself on EDGAR rather than a summary. The summaries reliably miss at least one of the five numbers below, and in 2026 the omitted number has repeatedly been the important one.
The five numbers, in reading order
| # | What to find | Why it is first |
|---|---|---|
| 1 | Coins bought or sold, and the date range | The sign of the number is the entire headline. The date range tells you which week you are looking at — often not the one just ended. |
| 2 | Average execution price | Compare it to the week’s spot range. Well outside the range means the disclosure covers a different period than you assumed. |
| 3 | Total holdings and average cost basis | The cost basis is the number that decides whether sales book gains or losses, and whether the company is underwater. |
| 4 | Securities issued (ATM programmes, preferreds, converts, bonds) | This is the funding side. Coins bought without new issuance came from cash; coins bought with issuance came from shareholders. |
| 5 | Use of proceeds — dividends, buybacks, general purposes | The tell. Proceeds from coin sales going to dividends is a different company than proceeds from share issuance going to coins. |
Work Strategy’s August 3, 2026 filing through the list as a live example. (1) 1,638 BTC sold. (2) Average $63,957. (3) 842,138 BTC held as of August 2, aggregate purchase price $63.51 billion, average purchase price $75,419 — so the sale was executed roughly 15% below cost basis. (4) Preferred stock outstanding, with STRC paying a 12.00% annualised dividend in semi-monthly instalments — a rate raised from 11.5% and a cadence moved from monthly to semi-monthly with effect from July 2026. (5) $52.4 million of the proceeds funded preferred dividends; $52.3 million funded repurchases of STRC stock. Five numbers, one conclusion: the company sold bitcoin at a loss to pay the coupon on the securities it issued to buy bitcoin. No commentary required; the filing says it.
Now run the same list over the following week’s August 10 filing and notice what changes. (1) 1,690 BTC sold, August 3–9. (2) Average $64,262. (3) 840,447 BTC, aggregate purchase price $63.36 billion, average cost $75,385. (4) Here is the number a headline reader misses entirely: 6,585,682 common shares sold for $653.1 million net. (5) Coin-sale proceeds went entirely to STRC repurchases. Read numbers one and five alone and you have a distressed seller. Read number four alongside them and you have something more interesting: a company that raised $653 million from equity markets in the same week it sold $109 million of coins — which means the coin sale was a choice about which pocket to fund the preferred coupon from, not evidence that the other pockets were empty. This is exactly why the funding line is number four on the list and not an afterthought.
The two arithmetic checks
Check one: reconcile the holdings. Take last week’s stated holdings, add or subtract this week’s coin movement, and see whether you land on this week’s stated total. If you do not, either you have missed a filing or a number is wrong — and it is almost always the first. A concrete case from this month: weekend coverage put Strategy’s holdings at 840,447 BTC while our own earlier reporting had 842,138. The gap looked like a discrepancy. It was not: 842,138 minus the 1,690 coins disclosed in the August 10 filing is 840,448, a one-coin rounding difference. The reconciliation resolved an apparent contradiction in about thirty seconds and required nothing but subtraction.
Check two: cost basis versus spot. Multiply holdings by the current bitcoin price, compare to the aggregate purchase price, and you have the unrealised position. At 840,447 coins and $63,600, Strategy’s stack is worth roughly $53.4 billion against an aggregate purchase price of $63.36 billion — about $10 billion underwater, or 16% below cost at an average basis of $75,385. That arithmetic is why the sales are happening at a loss, and it is the single most useful thirty seconds you can spend on one of these filings. Do it every week; the number moves.
mNAV: the ratio that decides which lever the company can pull
Market-to-net-asset-value — mNAV — is the company’s market capitalisation divided by the market value of its bitcoin. It is the most important number that is not in the filing, and you have to compute it yourself. Above 1.0, issuing shares to buy bitcoin adds bitcoin per share: the company sells a dollar of stock for more than a dollar of coin, and existing holders end up with more bitcoin behind each share. That is the whole trick, and it is genuinely accretive while it lasts.
Below 1.0, the trick inverts. Issuing shares now destroys bitcoin per share, so the equity-issuance lever closes and management must find another one: cash, debt, preferred stock, or selling coins. This is not theoretical in 2026. A sub-1.0 mNAV effectively suspended Metaplanet’s common-share issuance during the second quarter, and on August 13 the company launched “BitBonds” — its 21st through 24th bond series, roughly ¥200 million or $1.3 million, unsecured and unrated, carrying coupons of 4.0–4.3% over about three-year maturities, explicitly not principal-protected and transfer-restricted, distributed through its own licensed securities arm under Japanese private-placement rules, per CoinDesk. The size is trivial against a company holding roughly 43,000 BTC and targeting 100,000. The signal is not: it is a treasury company testing whether a domestic credit market will fund what its own shareholders will no longer fund at an accretive price. When you see a treasury company open a new funding channel, check its mNAV first; the new channel usually exists because the old one closed.
The four traps
- The date-range trap. The filing date is not the transaction date. Strategy’s August 10 filing covered trades executed August 3–9. If you compare the disclosed average price to the price on the filing date, you will conclude the company executed badly when it simply traded a different week. Always read the range, then compare to that week’s range.
- The headline-holdings trap. Aggregate holdings can rise while the company is a net seller in the period, if a purchase settled inside the same window. Read the movement line, not just the total.
- The use-of-proceeds trap. “General corporate purposes” and “to fund dividends on our preferred stock” are radically different disclosures. The first is a company with options; the second is a company with an obligation. Preferred dividends are contractual and recur on a schedule — Strategy’s STRC pays semi-monthly at a 12.00% annual rate — which means the funding requirement returns whether or not the bitcoin price cooperates.
- The framework trap. Corporate authorisations are ceilings, not forecasts. Strategy’s board adopted a BTC Monetization Program on June 29, under its Digital Credit Capital Framework, permitting up to $1.25 billion of tactical sales alongside a $2 billion buyback authorisation. That is a limit on what management may do, not a prediction of what it will do — but it does tell you the tool exists and has been pre-approved, which is why the second and third sales are less surprising than the first.
What the weekly filing tells you about bitcoin itself
For most of the last two years, treasury companies were a source of persistent, price-insensitive demand: they issued paper and bought coins on a schedule. Reading the 2026 filings in sequence shows that relationship inverting for at least part of the sector. Strategy sold 1,638 coins in the week to August 2 and 1,690 in the week to August 9, roughly $213 million of supply across two weeks, or about $15 million a day. For scale, the network issues roughly 450 BTC a day — around $29 million at current prices. Two weeks of one company’s sales was therefore equivalent to roughly half of new issuance over the same period.
That is a real number, and it is also a manageable one: US spot ETFs have absorbed multiples of it on single good days, and gave back $389.7 million in the week to August 14 alone. The point of reading the filings is not to find a single number that explains the price. It is to know, each week, whether a category of buyer that used to be structural has become a seller — and to know it from the primary document on the Monday it is filed, rather than from a summary on Thursday.
Five rules
- Read the primary filing. EDGAR full-text search, company ticker, form type 8-K. Two minutes. Summaries drop numbers.
- Reconcile before you react. Last week’s holdings plus this week’s movement equals this week’s total, or you are missing a filing.
- Read use-of-proceeds before you read the coin count. Where the money went tells you what kind of company you own.
- Compute mNAV yourself, weekly. It is not in the filing and it determines which funding levers are available.
- Match the date range to the price range. Every apparent execution scandal we have checked this year dissolved on this step.
What is a weekly 8-K for a bitcoin treasury company?
Form 8-K is the SEC’s current-report form for material events between quarterly reports. Bitcoin treasury companies such as Strategy use it as a near-weekly operating bulletin disclosing coins bought or sold, securities issued, dividends declared and use of proceeds. It is free to read on EDGAR.
What is mNAV and why does it matter?
mNAV is market capitalisation divided by the market value of the company’s bitcoin. Above 1.0, issuing shares to buy bitcoin increases bitcoin per share and is accretive. Below 1.0 it is dilutive, which closes the equity-issuance lever and pushes management toward debt, preferred stock or coin sales — as happened at Metaplanet before its August 2026 BitBonds launch.
How do I check whether a treasury company is underwater?
Multiply stated holdings by the current bitcoin price and compare to the aggregate purchase price disclosed in the filing. Strategy’s roughly 840,447 coins at about $63,600 are worth near $53.4 billion against an aggregate purchase price of $63.51 billion and an average cost of $75,419.
Why does the filing date differ from the transaction date?
The filing reports a completed period. Strategy’s August 10 filing covered trades executed August 3–9. Comparing the disclosed average price to the price on the filing date will produce a false conclusion about execution quality.
Do treasury company sales move the bitcoin price?
They are one supply input among many. Strategy sold roughly $213 million of bitcoin across the two weeks to August 9, about $15 million a day, against daily network issuance of roughly 450 BTC or $29 million. Meaningful, but smaller than a single heavy ETF flow day.
Disclaimer: This article is for informational purposes only and does not constitute investment advice. Cryptocurrencies and crypto-linked equities are volatile and you can lose money. Do your own research and consult a licensed financial advisor before making investment decisions.