Why This Matters
If you rely on quarterly crypto earnings to judge a firm’s treasury performance, you may be mistaking accounting noise for real profit or loss. Understanding the split between unrealized and realized figures helps you spot when a company’s bitcoin sales are actually adding value — or draining it.
Strategy Inc. reported an $8.32 billion unrealized loss on digital assets for the quarter ended June 30 2026, yet its earnings release showed only a $900 000 realized loss (CoinDesk). The gap stems from the difference between fair‑value remeasurement and the cost‑basis of the bitcoin actually sold.
The Gap Between Quarterly Income Statement Realized Losses and Annual Cost‑Basis Disclosures
Strategy’s July 6 2026 pre‑announcement told investors to expect an $8.32 billion loss on digital assets, broken into $8.31 billion unrealized and $900 000 realized (CoinDesk). When the results arrived on July 30 2026, the press release described the entire $8.32 billion as an unrealized loss, and the realized figure disappeared into a rounding convention (CoinDesk). This shift is not an error; it reflects how quarterly income statements capture only the change in fair value since the last measurement, not the total economics of a disposition.
The realized loss that did appear — $900 000 — represents the difference between the last fair‑value mark before sale and the proceeds, not the profit or loss relative to the original purchase price (CoinDesk). Because ASU 2023‑08 requires crypto assets to be carried at fair value with remeasurement through net income, the income statement effect is usually small and sometimes zero (CoinDesk). Investors who read the $900 000 as evidence of a minor loss are therefore looking at the wrong number.
The true economic impact of Strategy’s bitcoin sales emerges only when you compare the sale proceeds to the cost basis of the coins disposed — a figure that FASB requires only in annual footnotes (CoinDesk). For the 1 363 bitcoin sold on June 29‑30 2026, proceeds were $80.8 million at an average price of $59 256 per coin (CoinDesk). Against Strategy’s disclosed blended purchase price of $75 578 per coin as of June 30 2026, those coins cost roughly $103 million to acquire, implying a $22 million loss on that single disposal (CoinDesk).
How ASU 2023‑08 Shifts Crypto Accounting but Leaves Realized Gains Ambiguous
ASU 2023‑08, issued by the Financial Accounting Standards Board, mandates that in‑scope crypto assets be measured at fair value with changes flowing through net income (CoinDesk). The Basis for Conclusions notes that some practitioners treat the difference between the last mark and the sale proceeds as a realized gain or loss, but the Board explicitly said this does not represent the total realized gain or loss from disposition (CoinDesk). The Board anticipated exactly the confusion seen in Strategy’s reporting.
Because every unit is already marked to market under fair value, the cost‑basis election now drives only a disclosure requirement and has no effect on the income statement (CoinDesk). This is why the cost‑basis method attracts little scrutiny despite being the key to understanding the true profit or loss on a sale. The standard requires firms to disclose which cost method they use — FIFO, specific identification, average cost, or another — but does not mandate its use for quarterly earnings (CoinDesk).
Consequently, two firms selling bitcoin at the same price can report opposite realized outcomes simply by choosing different cost‑basis assumptions (CoinDesk). The accounting rule creates a scenario where the income statement tells you nothing about the actual cash‑flow impact of a crypto disposition, while the annual footnote holds the economically relevant data.
Strategy’s Bitcoin Sales Illustrate Why Cost‑Basis Election Matters
On June 29‑30 2026, Strategy sold 1 363 bitcoin for $80.8 million, an average of $59 256 per coin (CoinDesk). Using the blended average purchase price disclosed for its holdings as of June 30 2026 ($75 578 per coin), the implied cost basis was about $103 million, yielding a roughly $22 million loss (CoinDesk). However, if the firm had applied FIFO to a position accumulated since 2020 — when bitcoin traded in five figures — the same coins would have carried a much lower basis, potentially turning the sale into a gain.
Strategy’s May 26‑31 2026 sale of 32 bitcoin at an average price of $77 135 illustrates the point even more starkly (CoinDesk). Against the blended average price of $75 699 disclosed at that time, an average‑cost method would show a realized gain of roughly $46 000 (CoinDesk). Yet management disclosed on the earnings call that the cost basis of those specific coins was $125 464 apiece — about $4 million of basis against $2.5 million of proceeds, resulting in a loss (CoinDesk). The same transaction, same market price, opposite realized outcome depending solely on which coins the firm said it sold.
Because the cost‑basis election is disclosed only once a year in a footnote, investors cannot tell from quarterly releases which method is being applied (CoinDesk). This opacity allows a firm to manage the appearance of realized performance without changing its underlying bitcoin holdings.
Regulatory and On‑Chain Transparency Gaps Leave Investors Guessing
On‑chain data can verify the timing and quantity of bitcoin moved from Strategy’s known addresses, but it cannot reveal the firm’s internal cost‑basis assumptions (CoinDesk). The blockchain shows that 1 363 bitcoin left Strategy’s treasury on June 29‑30 2026, matching the disclosed sale size, yet it does not show what price the firm originally paid for those specific coins (CoinDesk). Without access to the private ledger, analysts must rely on the company’s disclosures.
The SEC requires firms to disclose their crypto accounting policies in annual 10‑K filings, but quarterly 10‑Qs need only repeat the income‑statement impact (CoinDesk). This creates a reporting lag: a change in cost‑basis method would not be visible until the next annual filing, potentially up to twelve months later (CoinDesk). For a fast‑moving asset like bitcoin, that lag can mask material shifts in treasury strategy.
Regulators have not yet mandated real‑time reconciliation of fair‑value marks with cost‑basis disclosures for crypto assets (CoinDesk). Until such rules appear, investors looking at quarterly earnings must treat the realized line item as a supplemental, not definitive, measure of disposition performance.
What Advisors Should Look for in Future Crypto Earnings
When reviewing a crypto‑heavy company’s quarterly report, advisors should first locate the unrealized gain/loss line and note its magnitude relative to the firm’s bitcoin holdings (CoinDesk). A large unrealized swing often reflects market moves rather than trading activity.
Next, check the footnotes for the annual cost‑basis disclosure and compare it to the quarterly proceeds from any disclosed bitcoin sales (CoinDesk). If the company reports sales but does not break out the cost basis, the realized figure in the income statement is likely insufficient to judge profitability.
Finally, watch for any changes in the cost‑basis method disclosed in the annual filing — switches from average cost to FIFO, for example — and assess how such a change could retroactively alter the perceived profitability of past sales (CoinDesk). Only by combining the income statement, the footnote disclosures, and on‑chain transaction verification can an advisor form a clear picture of a crypto treasury’s true performance.
| Bull Case | Bear Case |
|---|---|
| Strategy’s disciplined treasury‑as‑a‑business model could generate steady realized gains if it adopts a low‑basis cost method and continues to sell bitcoin at higher prices. | Reliance on quarterly unrealized figures may obscure deteriorating economics, leading investors to overestimate the profitability of Strategy’s bitcoin holdings. |
How should investors adjust their evaluation of crypto‑heavy firms when quarterly earnings hide the true cost basis of asset sales?
Key Terms
- Unrealized loss — a loss in value that exists on paper because an asset’s market price has fallen below its carrying amount, but the asset has not been sold.
- Realized loss — the actual loss incurred when an asset is sold for less than its original cost basis.
- Cost basis — the original purchase price of an asset, adjusted for any improvements or specific accounting elections, used to calculate gain or loss upon sale.
- ASU 2023‑08 — the Accounting Standards Update that requires crypto assets to be measured at fair value with changes flowing through net income.
- FIFO (first‑in‑first‑out) — a cost‑basis method that assumes the oldest assets are sold first, often resulting in lower basis for long‑held positions.
