The latest company-by-company coverage register records the reviewed live dataset, original disclosures, field checks and unresolved items. Download the full evidence and dated corrections there. Historical reviews below retain their original scope and dates.
This editorial and layout review covers Altius, Berkshire, BAM, Pershing, ASML, Alibaba, Netflix, Meta, Apple, Tortilla, Evolution and Valaris / Transocean. The later updates organise business explanations, core scorecards, specialist detail and reading paths; they do not establish a new filing-verification date.
Read dates separately: the editorial update date describes wording and structure. Each financial observation retains its reporting period and source; a page update does not make an older figure current. Historical reviews below describe their original coverage.
Core metrics provide an introduction. Detailed metric groups, exact-value tables, disclosure gaps and downloads remain available. Calculation checks do not constitute an independent audit of company accounts.
The site presents historical NAV growth, financial results and explanations of business operations. Forecasts, valuation models, projected shareholder returns and hypothetical financial calculators have been removed, including forward-looking values in downloadable company datasets. Historical results and saved research notes are preserved. The site does not recommend stocks.
This editorial update does not refresh source verification or change access. Earlier model-related review notes below describe the previous version.
The new Apple page has a separate source and calculation review: FY2021–2025 plus Q3 and nine-month FY2025–2026 comparisons. Category sales, gross profits, margins and cash-flow calculations are reconciled to Apple filings and releases. Device and subscription milestones remain separately dated; no continuous retention, replacement-cycle or services operating-margin series is asserted.
This addition brings the centre to 12 company pages. The earlier whole-site review below covered 11. Browser visual QA was unavailable; automated calculation and interaction checks were used for Apple.
Whole-site data review · 19 October 2026
Reviewed all 11 company pages and their portfolio sections for missing figures, source mismatches, calculations and reporting-basis problems. The corrections below are now applied. Coverage varies by field; the table distinguishes fresh checks from earlier source work and unresolved gaps.
Pershing’s FY2021 Netflix company FCF is now −$158.894m. The separate CFO less PP&E proxy remains −$131.975m. The $26.919m other-assets cash outflow explains the difference.
Evidence ↗ · MD&A FCF reconciliation, printed p.27
Four Berkshire holdings and nine Pershing names had vendor caps that exactly implied diluted average EPS shares; Microsoft’s vendor denominator was also unestablished. Berkshire now uses explicitly dated common-share value proxies. Ten additional Pershing raw caps are withheld; Microsoft, Amazon and S&P Global have separately labelled June-share estimates, and QSR retains its sourced economic-share calculation. Unsupported cap-dependent ratios are unavailable. Raw vendor observations and dated prices are preserved.
Evidence ↗ · Per-company evidence and raw observations
The broad total-catastrophe-loss field is now unavailable instead of zero. A separate flag records that no event exceeded the reported $150m pre-tax threshold.
Evidence ↗ · PDF pp.17 and 35
Added ASML High/Low NA recognition and a cumulative shipment milestone; Meta FX-neutral ad growth; Tortilla leases, franchise income, store split and LFL; Evolution availability and half-year geography; portfolio closing shares, book values, Microsoft cash/debt and ICE buyback zeros; and Valaris split-period financing and closing-share observations. Each field includes its source and calculation or reporting basis.
Evidence ↗ · Detailed changes and evidence
The fee-bearing capital bridge now explains that reported inflows already include deployment of previously uncalled commitments. Deployment must not be added again. The separate annual margin table now also includes its sourced 2021 value.
Evidence ↗ · Printed pp.6 and 28
ASML quarterly annual-sales coverage is labelled not applicable; Q3 investment gains and downpayment effects are explained. Coca-Cola’s prior adjusted-FCF source was added. Newly filled fields no longer carry stale missing-data notes.
Evidence ↗ · Dated sources and calculation inputs
| Company | Checked and updated | Remaining limits |
|---|---|---|
| Altius | Annual 2021–2025 and current interim metrics, six latest snapshots, portfolio carrying values, bridge and selected event figures were checked against primary reports. No new numerical correction found. | Earlier royalty-definition breaks and the conflicting 2021 MD&A comparative remain explicit. |
| Berkshire | Annual and interim financial/insurance data and all five holdings’ historical financial inputs were reconciled; 104 independent arithmetic checks passed. Share-count evidence now supports dated equity-value proxies. | The holdings financials remain FY2025 snapshots. Prices are dated vendor observations; actual shares are from separately dated filings. Alphabet’s full D&A remains unavailable. |
| PSPSHPSUS | All 16 operating-company scorecards, fund data and all 15 manager-wide 13F rows were covered. Fresh arithmetic, latest-source and selected historical checks supplement the 18 October review; 177 arithmetic checks passed. | This is not a fresh extraction of every historical field. Missing share-class denominators, lease/software components and S&P Global perimeter history remain qualified. Forward-looking manager estimates were subsequently removed under the historical-information policy. |
| ASML | Annual and quarterly financials, product units, bookings and backlog were checked, including Q3 2026. 110 arithmetic checks passed. Five missing EUV recognition/shipment observations were added. | Physical shipment counts differ from revenue recognition. Quarterly bookings are no longer disclosed. Backlog/annual-sales coverage is not applicable to a quarter. |
| Brookfield BAM | The 29 annual observations prepared on 19 October were retained; latest supplement figures, strategy totals, operating leverage, Oaktree comparison and model arithmetic were rechecked. | 2021 fee-eligible uncalled commitments remain unresolved (approximately $40bn versus an implied $44bn). Pre-listing scope, share denominators and large insurance mandates limit comparisons. |
| Alibaba | The 208 sourced historical observations prepared earlier on 19 October were retained. All 76 reporting-basis/period selections, margins, FCF reconciliations and original June-quarter comparisons were retested. | Current combined-commerce and Cloud/AI scope cannot be projected backwards across all five years. Earlier variants remain separate; missing current-scope annual figures are not zero. |
| Netflix | Annual 2021–2025 and Q2/H1 2025–2026 financials, content cash flow and per-share calculations were checked. The dedicated page’s 2021 FCF was already correct; the duplicate Pershing field was corrected. | Exact membership and ARM reporting stopped in 2025. Membership milestones are bounds; exceptional receipts prevent a simple normalized-FCF claim. Q3 is scheduled after this review date. |
| Meta | Annual and interim financials, segment results and cash-flow formulas were checked alongside Netflix (81 combined formula checks). Seven missing constant-currency advertising-growth figures were added. | Reality Labs capex, platform ROAS, standalone advertising FCF and comparable early DAP-based ARPP remain undisclosed. Q3 is scheduled after this review date. |
| Tortilla | Core cash-flow, debt, earnings, store and restated interim values were checked. Twelve fields were added, including leases, franchise income, the 2024 estate split and selected LFL figures. | FY2022 company-operated store prose conflicts; H1 2025 estate timing is unclear. Full French closure cash costs and kitchen/store unit economics remain gaps. |
| Evolution | Core financials, FCF, shares, operating counts and regional definitions were checked. Seventeen fields were added: availability and calculated H1 regional/regulated measures. Tortilla/Evolution together passed 95 arithmetic checks. | The 2021 game count and later-definition Latin America split remain unavailable. Utilisation, studio ROIC and new-game revenue share are not inferred. |
| Valaris / Transocean | Core financials, debt/cash definitions, financing flows and current merger status were reviewed; 82 arithmetic checks passed. Six source observations fill eight displayed cells, including the 2022 closing-share change. | The review did not re-extract every operating-statistics or merger-projection row. Fresh-start periods stay separate. Interim cash interest, maintenance-only capex and realised merger savings remain unavailable. |
Examples: Reality Labs capex; advertising ROAS; Tortilla kitchen utilisation and full French closure cash costs; Evolution studio utilisation/ROIC; maintenance-only offshore rig capex.
Examples: Alibaba’s current segment perimeter before its recast; Tortilla France before acquisition; Valaris full-year 2021 cash flows across fresh-start accounting; ASML quarterly coverage using an annual-sales denominator.
Examples: BAM 2021 uncalled fee runway; Tortilla 2022 own-store count; Altius’s original 2021 MD&A comparative. Conflicts remain documented rather than resolved by guessing.
Examples: some portfolio share-class, lease and software histories; some operating-statistics and merger-projection rows. Existing source provenance is retained, with bounded review coverage above.
Market prices remain 16 October snapshots. Period labels govern financials; the review does not certify that every company uses its newest available release. The current site does not display forecast results or scenario outputs.
Missing values are never treated as zero. Verified zeros, such as ICE’s suspended programme repurchases, carry explicit evidence. Calculated values retain their formulas and dated inputs.
Automated calculation, rendering and data checks were used. Browser visual QA was unavailable. This review is dated and is not a blanket guarantee of accuracy.
Archived findings and coverage below describe the earlier review. The 19 October corrections above take precedence, including the updated market-cap denominator findings. Earlier raw vendor evidence remains available for traceability.
Reviewed Altius, Berkshire, Pershing Square Holdings, ASML, Brookfield Asset Management and Alibaba, including metric explanations, company scorecards and scenario output text. That review retained hypothetical models with qualified wording. The 19 October editorial update subsequently removed those models. No buy, hold or sell recommendations are provided.
The later update preserves navigation, historical financial data and source references, and removes hypothetical inputs and model functions. User-written research notes and their saved versions are preserved. This is an editorial scope review, not legal certification or a new verification of every financial figure.
The Netflix page adds 2021–2025 annual financials and matching Q2/H1 2025–2026 comparisons from primary reports. Membership milestones, advertising bounds, the 2021 FCF adjustment, share-split basis and 2026 exceptional cash receipts are qualified. This is separate from the earlier site-wide review and does not extend its metric ledger.
Use later economic-ownership comparative; retain original amount separately. 2021–22 remain on original basis and are not joined to 2023–25 in the trend line. Growth across the definition break is suppressed.
Primary evidence ↗ · PDF pp.10,29
Replaced hardcoded growth with 2020 common equity305.816 / shares41.477653m as sourced baseline. No change to rounded displayed growth.
Primary evidence ↗ · PDF pp.6,10
Explicitly excludes investing-classified JV distributions as well as stated acquisitions, growth capital and financing payments. 2021–22 royalty figures remain as originally reported, not an invented modern-basis restatement.
Direct reported totals and full-precision average shares now cite earnings release B6 rather than component/rounded tables. Operating earnings44,486m and47,437m unchanged.
Primary evidence ↗ · PDF pp.1–2
Capital citation corrected from printed p.14 to p.4. The10.5% value is unchanged.
Primary evidence ↗ · Printed p.4
Latest pro forma comparable is2,937. TTM=6,140+3,684−2,937=6,887. Pro forma figures remain labelled separately from historical reported GAAP.
Primary evidence ↗ · Exhibit4
Vendor cap/price implies1,643.737m shares, materially inconsistent with the share-count scale in split-adjusted issuer reporting. No exact replacement inferred from rounded diluted averages. Cap-dependent ratios are unavailable; sourced per-share DE and price metrics remain.
Primary evidence ↗ · Operating results and split notes
SourceC14 now points to parent-company10-K rather than a partnership filing that did not support all denominator inputs.
Primary evidence ↗ · Equity statement and EPS notes
Headline now identifies a proxy: operating income plus aggregated cash-flow D&A. It is not company-reported/adjusted EBITDA; the aggregate can contain financing-related amortisation.
2024–25 High NA counts are derived from cumulative shipment milestones, not directly reported annual customer-acceptance totals. First modules can ship before installation. FY2021 backlog≈€24.2bn is derived from rounded2022 €40.4bn and67% growth, not directly reported precise backlog.
US Treasury bills separated from treasury stock and preferred/senior claims; growth, price/book and debt/EBITDA explanations clarified. Average margin no longer described as incremental margin.
All stored prices/caps reproduced in the web finance feed with16October timestamps. Raw returned evidence and query identifiers are retained in market-data-audit.json. This is not independent primary validation of the feed, and market-cap share counts remain vendor assumptions.
Azure metric includes other cloud services; Visa cross-border growth uses constant dollars; Mastercard GDV includes purchases and cash activity. QSR compensation includes non-cash incentives, with calculated TTM140m versus issuer-rounded139m disclosed. Meta40976 is a gross-capex proxy, while company FCF after lease principal is37872. The return model examined in that review has since been removed.
Newer five-year table in Q4 FY2025 investor presentation reports20,041; original FY2023 release20,040 is retained in the dataset as an original observation. Book-to-bill recalculated.
Primary evidence ↗ · Five-year financial summary
88 metric cells,24 derived BV/FCF/net-cash checks, latest snapshots,13 portfolio rows,7 potash subrows,8 bridge values and7 selected events reviewed against filings. Corrections/qualifications above apply.
Annual2021–25 and2026 interim financial/insurance values, segment reconciliations, top-five holdings financial inputs and cited management disclosures reviewed. Quote observations have separate vendor status.
254 present metric-period values inventoried. 100 arithmetic identities recomputed, raw financials compared with annual/quarterly statements, unit charts and shipment remarks. Missing/derived milestones retain explicit status.
Completed context review of populated raw financials, annual histories, YTD columns, balance sheets, dilution, business KPIs and cash adjustments for all16 upgraded issuers, plus matching legacy cards, PSH fund data, all15 SEC13F rows and38 manager-estimate inputs. 180 derived numeric fields were independently recalculated in the portfolio review. Those manager forecasts were subsequently removed. Field-level inventory and status distinctions are downloadable.
Annual financial histories and 349 business-metric observations now cover the five-year windows of all16 operating-company scorecards. 41 unavailable or incompatible business observations retain explicit reasons. S&P Global’s original group is separate from its current pro forma perimeter. Prices, portfolio weights and valuation snapshots are not retrospectively invented.
Source and calculation review. Not an independent professional audit opinion. Verification is dated; forecasts and investment interpretations are not reported facts.