Corrections

The claims assembly read an absent input as a zero. A blank cell and a filed zero are different facts: a filed 0 says the company has no debt, and a blank says nobody has filed the debt yet. The shared reader coerced both with the same expression, so a row awaiting a figure contributed nothing to Net Senior Claims and rendered as though the figure were known to be nil. Capital B is the instance. Its 2026-09-07 row carried blank debt_usd, blank cash_usd and a blank btc_price alongside the post-split re-key, and the coercion dropped $4,775,026 of debt against $4,029,310 of cash, the two figures the 2026-09-02 row carries. Senior Claims % served 22.15 where the same row with both cells filed gives 22.42, and CEBE served 7,166 sats where it gives about 7,142. Both errors run in the direction that flatters the issuer. The reader now withholds the figure and names the missing cell rather than computing across it, and the surfaces render "n/a, row incomplete" with the reason. A filed zero is unaffected and still computes, as are the two columns whose blank genuinely means none, strc_holdings_usd and btc_claim_static, each of which has carried its own absent-is-not-zero reader since an earlier incident. The same admission test refuses a row with no usable Bitcoin price, which previously dropped the entire fiat leg and served the BTC-denominated claim alone as though it were the whole stack.

Source: Snapshots tab, ALCPB_20260907_R (btc_holdings 3,521; debt_usd blank; preferred_usd 0; cash_usd blank; btc_claim_static 779.95; btc_price blank) against ALCPB_20260902_R (debt_usd 4,775,026; cash_usd 4,029,310), both as served 2026-09-08.

One row of the 87 on the tab trips the rule today and it is Capital B's latest, so Capital B is the only company whose served figures move. Served wrong from 2026-09-08, bounded by observation rather than asserted: /api/v1 served the 2026-09-02 row at 11:35 UTC and the blank 2026-09-07 row at 12:09 UTC. The window closes when the corrected reader ships with this entry. No other company's figures change, and no historical row is affected.

STRK's liquidation preference was described as an upward ratchet on The Net Effect and as $1,000 per share in the glossary. Under the amended Certificate of Designations it is the greatest of the $100 stated amount, the last reported sale price and a ten-day average of those prices, so it floats daily above a $100 floor and can fall as well as rise. STRF, STRC and STRD carry a $100 stated amount. Both passages are corrected to the certificate's terms and The Net Effect's section heading follows.

Source: STRK Certificate of Designations as amended July 2025, as cited on /the-net-effect/ section 04; the Strategy 10-Q guide at /guides/reading-10q-treasury-numbers/ already carried the correct reading.

Effect on served figures: none. preferred_usd already carries the accrued liquidation preference as filed.

Two definitions of break-even were live. The glossary and the 2026-07-16 entry below stated Net Senior Claims (USD) over Total BTC; the comparison table, the API and the company page solved the fiat leg only, net fiat senior claims (USD) over Bitcoin held net of BTC-denominated claims, null where the fiat claims net to cash or the indexed claim covers the treasury. The served surfaces were right and the document of record was wrong: the glossary form is circular for a BTC-indexed issuer because Net Senior Claims USD is itself a function of the price being solved for, and it coincides with the served form only when the static claim is zero. For Capital B at the 2026-09-02 row (3,145 BTC, cash $4,029,310, 821 BTC static, btc_price 77,297) the glossary form gives $18,897 and the ruled form gives no break-even, net cash. The 2026-07-16 entry installed the wrong form and is annotated below. The glossary, the drawdown template (which also ignored the static term) and the modeler engine are corrected.

Source: Ruling R-0906 F-18, restated R-0907 F-4; reference implementation compare/compare.js and api/v1/history.js.

Effect on served figures: the comparison table, the API and the company page are unchanged. The drawdown pages and the modeler print n/a with the reason for net-cash and indexed-covered issuers where they printed $0 or a dash; no live drawdown page carries such an issuer today.

The company page's amplification footnote asserted 1 / (1 minus Senior Claims %) and its code used the floored share, while the ratified definition of 2026-07-25 is (1 minus Indexed Claims %) over (1 minus Senior Claims %), unfloored, carried in config/compute-amplification.js and used by nine other surfaces. For Capital B at the 2026-09-02 row (3,145 BTC, 821 BTC of BTC-denominated claims, net cash) the page printed 1.32x where the ratified value is 0.98x. The page now reads the shared module and the footnote states the ratified form.

Source: config/compute-amplification.js, ratified 2026-07-25; R-0907 F-3.

Served wrong on /company/ALCPB/ from the key block's launch on 2026-08-28 to 2026-09-07. Issuers with no BTC-indexed claim printed the same figure under both forms and are unaffected.

The company page and the tracker re-marked Strategy's EUR-denominated STRE preferred (EUR 775,000,000) from the snapshot's filed cross to the live EUR cross before the claims term, while the API, the comparison table, the charts explorer and the drawdown pages carried preferred_usd as filed. One company and one date therefore had two CEBE figures on one site, and the band chart's caption promised a snapshot basis the chart did not have. Ruled: a non-USD preferred face enters every surface at the snapshot's filed cross, never re-marked. On the 2026-08-30 row (preferred_usd 14,789,035,800 at eurusd_rate 1.1408) the company page printed 156,719 sats and a $17,482 break-even at a live print of $79,359 where the filed basis gives 156,769 sats and $17,462; at the snapshot print the band chart printed 155,843 where every other surface printed 155,895. The tracker carried the re-mark in its USD net claims figure and not in its CEBE cell, so one metrics object disagreed with itself.

Source: Snapshots tab, MSTR 2026-08-30 row (preferred_usd 14,789,035,800; stre_face_eur 775,000,000; eurusd_rate 1.1408); R-0907 F-1.

Served wrong on /company/MSTR/ from 2026-05-11, when the re-mark landed, to 2026-09-07, 119 days; the tracker's modal USD net claims figure over the same window. Strategy's company-page CEBE moves by about 50 sats and its break-even by about $20 to agree with the other surfaces; 155,895 sats at the 2026-08-30 snapshot print is the figure.

The page attributed a 1.22x accretion threshold to Strategy’s Q1 2026 earnings disclosure. The figure sits in the earnings presentation cited below rather than in the earnings release, stated as of 2026-05-03, and was superseded as a construction on 2026-07-23 when Strategy redefined mNAV with a 1.0x threshold. Source note corrected and a dated paragraph added. Found by the MSTR anchor on a sourcing check dispatched 2026-09-02.

Source: Q1 2026 Earnings Presentation of 2026-05-05, slide 59.

Effect on served figures: none

DDC Enterprise's Senior Claims % was understated, on a debt figure of $31,323,579 where the correct figure for the same balance-sheet date is $85,291,473. The claims stack omitted a Bitcoin-collateralized borrowing of $50,980,369, being RMB 356,510,817 at the filing's 6.9931 convenience rate, secured on 601 Bitcoin pledged to third-party creditors. The issuer filed it under accrued expenses and other current liabilities rather than under borrowings, and the tracker's construction read the borrowings captions. A second and much smaller component moved the July 2025 senior secured convertible note from its amortized carrying value of $19,662,475 to its face of $22,650,000, bringing DDC into line with the standing rule that senior claims are carried at face. Corrected Senior Claims %: 2025-12-31 reads 100.00% against 58.98% served, 2026-03-31 reads 72.92% against 39.59%, and the running 2026-07-03 basis reads 35.95% against 6.19%. At the 2025-12-31 balance-sheet date the corrected senior claims exceed the Bitcoin treasury, so Claims % pins at 100.00% against an uncapped 111.07% and common equity Bitcoin exposure at that date is negative. The pledge carries a release gate at a loan-to-value ratio of 60 percent and no maintenance covenant behind it, so a decline in Bitcoin locks the collateral in rather than forcing a sale; nothing here describes a liquidation trigger. Two rulings now stand for this issuer: senior claims are carried at face and never at carrying value, and a senior claim counts wherever the issuer files it, so liability notes are read for borrowings rather than the borrowings caption alone.

Source: FY2025 Form 20-F, accession 0001213900-26-045881, filed 2026-04-21, Note 15 cross-referenced from Note 12 footnote (ii). Release condition per the SPA, accession 0001213900-25-063293.

Served wrong from 2026-04-21 to 2026-09-01, 133 days. The period runs from the date the correcting disclosure became public, not from the date the figure first became wrong.

Desk attribution corrected for the Barchart and AlphaQuery vendor captures behind the R-VOL companion's coverage table. The page named no desk for them, and the only capture actors it did name were tied to the Market Chameleon work, so the captures read as that desk's. They were performed by the Cowork desk on 2026-08-07 and delivered 2026-08-15. The Market Chameleon desk has attested zero vendor access, carried as NOT_CAPTURED_BY_THIS_DESK on every row of the R-VOL primary table. The coverage claim of 12, 6 and 2 dates of 12 stands on the same evidence, and the verdict is unchanged.

Effect on served figures: none

Strategy's Senior Claims % and CEBE were computed with declared-but-unpaid preferred dividends carried as a separate claims line. The ruling of record is that no standalone dividends-payable term exists in the claims stack. The accrued liquidation preference is the carrying value, and a payable is settlement in transit, not an additional claim. Corrected figures at the 2026-08-02 snapshot basis: Senior Claims % 33.71 (was 34.00), CEBE 144,023 sats (was 143,392). The dividends-payable figure remains disclosed as a non-netting field. No other issuer carried the term. Checked 2026-09-07: the 2026-08-02 row as served (btc_holdings 842,138; debt_usd 6,753,703,000; preferred_usd 15,341,948,700; cash_usd 4,073,420,776; shares_outstanding 387,605,620; btc_price 63,466; dividends_payable_usd 155,157,000) gives Senior Claims % 33.72 and CEBE 144,005 sats, 1 basis point and 18 sats from the figures above. The inputs this entry was struck on were not pinned, so whether the row moved after 2026-08-05 or the entry was struck on different inputs cannot be shown from the entry; the row's fields are pinned here and the row as served is the figure.

The five-term construction was live from 2026-07-30 to 2026-08-05

The Q2 2026 results refute the STRC stub-week finding. Wrong: the audit's STRC share count, 2,275,613 shares below the dashboard's, which reconciled to the consolidated preferred totals only at an implied EURUSD rate of 1.4345 against an actual 1.14133 on June 30. Right: the dashboard's count, which reconciles to the 8-K's consolidated balance sheet at $15,462,056 thousand of preferred redemption value and liquidation preference and 153,529 thousand preferred shares issued and outstanding. The stub-week shares were sold March 30 to 31 and issued in early April, and the audit read a settled-basis identity as proof they were already inside the March 31 base. The forward test staked in the piece resolved against its own prediction.

Source: 8-K, accession 0001628280-26-051027, furnished 2026-07-30.

Effect on served figures: none. The served figure was the one the piece described as inherited, and it was correct.

Correction and update log. The original 10.6x preferred-event ratio came from a 64-week synthetic dataset whose inputs could not be reproduced. The dataset was rebuilt from verified 8-K filings on 2026-04-27, the Q1 2026 debt and cash anchors were corrected on 2026-06-07, and the series was extended to 66 weeks on 2026-07-01.

Source: Strategy 8-K series January 2025 to June 2026 as reconstructed on /preferred-era/; dataset rebuilt 2026-04-27, Q1 2026 debt and cash anchors corrected 2026-06-07, series extended to 66 weeks on 2026-07-01.

Old: a 10.6x preferred-event ratio from a 64-week synthetic dataset whose inputs could not be reproduced. New: the ratio is retired and the finding is binary, that preferred events are the sole source of cumulative expansion in the gap between fully diluted BPS and CEBE for Strategy, tested on 66 weeks of filed data to 2026-07-01 with the Q2 and Q3 2026 checkpoints staked on /evidence/. Effect on served CEBE figures: none; the change is to the article's claim and its dataset.

Methodology page batch, five additions. An independence credential added to data standards, no CEBE figure depends on any issuer-defined metric; the cash netting rule stated beside the Net Senior Claims definition; the numéraire stated as a convention rather than a uniquely valid frame; a claims-scope note; and the accompanying basis, the criterion-7 sweep run 2026-08-02, verdict clean.

No computed construction changed

Two rulings land as page text. Ruling 2.1 names and holds apart three share-count constructions: basic shares outstanding, market-convention fully diluted, and the moneyness-partitioned construction an issuer published in July 2026. The comparison column stays on the all-dilutives convention deliberately, as the harsher of the two. Ruling 2.2, retirement conditionality, ships as a new crossover section on the retirement page.

No computed construction changed

Adjusted Claims % specification amended to v1.1 following an implementation review, with seven clarifications. None changes a computed construction.

Effect on served figures: none

Section 3.3's external-review credit corrected to Valentin Kosanovic. The surname was misattributed at initial publication. The credited review and the mark rule are unchanged.

Effect on served figures: none

Senior claims corrected, and the static BTC claim given one reader. A filing-true reported zero in btc_claim_static was read as an empty column and replaced by a fallback override, so a settled claim kept being counted.

Claims % 19.5% to 12.9%, CEBE 584 to 632 sats, at $66,077

Superseded 2026-09-07: the definition this entry installed, Net Senior Claims (USD) over Total BTC, is the wrong form for any issuer with a BTC-denominated claim; see the 2026-09-07 break-even entry above. Original text follows. Break-Even BTC Price definition corrected. The glossary had defined it through a market-parity construction. Break-Even is computed as Net Senior Claims (USD) / Total BTC, a balance-sheet level independent of the stock price.

Effect on served figures: none, the error was confined to the glossary

Zero floor moved from the fiat bucket to the total. The published formula and the serving code disagreed for issuers holding both BTC-indexed claims and surplus cash, because the code discarded the cash surplus.

Source: Snapshots tab, ALCPB 2026-06-01 row (btc_holdings 3,139; debt_usd 0; preferred_usd 0; cash_usd 8,610,000; btc_claim_static 1,011; btc_price 71,360; shares_outstanding 300,650,632), the last row before the correction; the adversarial audit of 2026-07-09, finding 3.

Capital B (ALCPB) Senior Claims % moved from 33.77% to roughly 29.4% at the print then live. Pinned 2026-09-07 on the 2026-06-01 row at its own price: the per-bucket floor gives 32.21% and 708 sats, the total floor gives 28.36% and 748 sats. The entry's own figures were struck when the static claim was carried at an override value since superseded by the filed 1,011, which is why they do not reproduce from the row as served.