Evidence layer for the whole dashboard
Uncertainty travels with the data
Every metric here has a source page, a verification status and, where sources disagree, a governed decision. This is a public product feature, not a hidden disclaimer.
bse_canonical_v1_0_2, database status approved,
approved_by = "Thegreat P. Malema" (approved 2026-09-08). The release was moved to
approved in Phase 9B, after the full bse stack was loaded to the
production database and independently verified there. The batch-05 / batch-06
structured disclosures behind Pages 5 & 6 passed the GATE A2 evidence review on
2026-09-07: 186 substantive rows visually verified against the source report, 13 control /
staging / manifest rows accepted as NOT_APPLICABLE, 0 findings.
bse_disc_release.status is now A2_VERIFIED (formerly
VISUAL_FIRST_PASS) and was not given a separate production approval beyond GATE A2.
Dashboard deployment and publication are governed separately, under GATE B (Phase 10-11).Reconciliation register at a glance
Phase A blind review and Phase B comparison are complete for all 26 cases. This case-level count is different from, and not comparable to, the individual-observation counts below: a single case (e.g. R007) can govern several observations.
Verification-status legend
| Status | Meaning |
|---|---|
| Verified | original PDF page, value, unit, period, entity and scope checked (per year) |
| Provisional | source calls it provisional / placeholder, or original-year coverage is outside the project; visual transcription does not remove this |
| Under review | two sources conflict, or scope / period / definition is unclear; withheld from headline cards |
| Derived | dashboard calculation: formula, inputs and rounding exposed; no more reliable than its least reliable input |
| Unavailable | not reported / not loaded / unsafe to calculate; shown as an explicit state, never zero |
| Forward-looking | targets and plans, never combined with achieved results |
Governed observations (individual data points)
Governed observations
Canonical (selected value)
Public (approved for display)
Provisional
Withheld from trend
Restricted display
"Withheld" (mart_withheld) and "restricted" (mart_restricted) are two separate governed
tables with different treatment: withheld observations are excluded from trends entirely;
restricted observations may still display with a scope/method label. They are not the same
population and are shown as two separate counts, not combined.
Canonical vs public: what the two views mean
Two different questions, and neither view contains the other:
| View | Rows | Means |
|---|---|---|
| Canonical | 247 | the single observation explicitly selected as the analytical value for a metric / period / scope |
| Public | 340 | every observation approved for public display (any PUBLISH* status) |
public − canonical = 93 rows: provisional historical comparatives for FY2018-FY2021. BSE
reports these pre-2022 points only in later five-year / comparative tables, so the dashboard
shows them (hollow / hatched, dashed connectors, provisional badge) but they are not canonical
selections. They are not rejected conflicts, restricted rows or duplicate keys.
Governed reconciliation register (R001-R026)
| ID | Canonical decision | Reason | Public treatment | Status |
|---|---|---|---|---|
| R001 | Use FY2024 P1,660,061,437.85 / P1,660.1m | Original-year detailed report outranks 2025 summary comparator | PUBLISH | Resolved |
| R002 | Use FY2024 P6.7m | Original-year detailed report corroborates repeated detailed tables | PUBLISH | Resolved |
| R003 | No canonical FY2025 velocity | 9.9% and 8.5% lack reconcilable definitions | WITHHOLD / SCORECARD-ONLY | Open / hypothesis |
| R004 | Use original-year reported 3.3%; retain 3.1% closing-cap calculation separately | Likely denominator timing difference | PUBLISH_WITH_WARNING | Resolved |
| R005 | No canonical FY2025 retail-share point | 15.1% conflicts with 4.5% | WITHHOLD | Open / hypothesis |
| R006 | Use FY2024 local-individual share 11.6% | Original-year report corroborates scorecard | PUBLISH_WITH_WARNING | Resolved |
| R007 | No single FY2025 debt total; retain 117, 121 and 122 with scopes | Debt-population figures refer to different source contexts and denominators; retaining scoped figures avoids a false merged total. | RESTRICTED_DISPLAY | Open / hypothesis |
| R008 | No canonical maturities count | 56 conflicts with 57 | WITHHOLD | Open / hypothesis |
| R009 | Use distinct issuer, equity-counter and product-inclusive metric IDs | Arithmetic is explainable when populations are named | PUBLISH_SCOPED | Resolved |
| R010 | Audited revenue and strategy revenue index are separate metrics | Audited revenue is a Pula financial-statement measure; the strategy revenue index is a separately defined scorecard construct with a different scope and denominator. | PUBLISH_SEPARATELY | Resolved |
| R011 | Retain each exact target label and horizon | 10% and 25% may describe different horizons/constructs | RESTRICTED_DISPLAY | Open / hypothesis |
| R012 | Primary ROE = PAT / average equity; closing-equity ROE is supporting | PAT divided by average opening and closing equity reproduces the reported 12.3%; closing-equity ROE remains a separately labelled supporting basis. | PUBLISH_WITH_METHOD_NOTE | Resolved |
| R013 | Original-year flags removed for verified 2022-2024 points | 2021 remains later-comparative provenance | MIXED | Partly resolved |
| R014 | Use restated FY2023 PAT P24,938,848 | 2024 audited restatement explains original P26,433,048 | PUBLISH | Resolved |
| R015 | Preserve published risk title plus working label | Duplicate heading remains | RESTRICTED_DISPLAY | Open / hypothesis |
| R016 | Keep committee populations and names separately scoped | Five/six and naming differences remain | RESTRICTED_DISPLAY | Open / hypothesis |
| R017 | Use visually located 3.6 only | Candidate 3.8 is unverified | PUBLISH_STANDALONE | Partly resolved |
| R018 | Use 30.3% from exact audited amounts | 30.4% arises from rounded headline figures | PUBLISH | Resolved |
| R019 | Audited cash totals publish; component relabel withheld | Cash bridge classification remains unclear | PARTIAL | Open / hypothesis |
| R020 | Keep composition and attendance populations separate | Seven vs eight may reflect timing/capacity | RESTRICTED_DISPLAY | Open / hypothesis |
| R021 | Use FY2022 velocity 2.9% | Original 2022 table and 2023 table agree; 4.9% prose is erroneous | PUBLISH_WITH_NOTE | Resolved |
| R022 | Assign 8.5% to FY2023 | Original 2023 report confirms period | PUBLISH_WITH_NOTE | Resolved |
| R023 | Use FY2022 closing cash resources P112,722,445; retain P112,722,304 as a conflicting opening balance | Four visual confirmations support P112,722,445. P112,722,304 appears once as an unexplained opening balance and is not labelled as a restatement. | PUBLISH_WITH_METHOD_NOTE | Resolved |
| R024 | Map 1,709.8 LASI to FY2023 | Narrative sequence and 2024 comparator corroborate period | PUBLISH_WITH_NOTE | Resolved |
| R025 | Preserve components; publish 0.9-point reconciliation gap as DQ note | Printed components sum 99.1% while total says 100.0% | PUBLISH_WITH_WARNING | Open / hypothesis |
| R026 | Use FY2022 DCI 7,726.4; retain 7,710.06 as conflicting later-summary evidence | The 2022 and 2023 Annual Reports agree at 7,726.4; the 2025 five-year table reports 7,710.06 without a reconciliation note. | PUBLISH_WITH_METHOD_NOTE | Resolved |
The cases carried on every affected chart
| ID | Metric | Governed outcome |
|---|---|---|
| R023 | FY2022 closing cash resources | Use P112,722,445. Retain P112,722,304 as a disclosed, immaterial P141 source inconsistency. |
| R026 | FY2022 DCI | Use 7,726.4 (2022 & 2023 Annual Reports agree). The 2025 five-year table's 7,710.06 is excluded from the trend and kept only in the evidence layer. |
| R025 | FY2024 investor mix | Printed components sum to 99.1%, not the printed 100.0%. Components are preserved, not renormalised; the 0.9-point gap is published as a data-quality note. |
| R003 / R005 | FY2025 velocity / retail share | No canonical value; withheld from ordinary trends. |
| R007 | Year-end listed debt instruments | 117 / 121 / 122 kept as separate scoped variants, never one total. |
Restatements
The FY2023 original PAT (P26,433,048) and restated PAT (P24,938,848) remain linked, not overwritten.
Historical availability
| Metric family | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|---|---|---|
| BSE Group financials | N/A | N/A | N/A | Partial | Available | Available | Available | Available |
| Counter performance | N/A | N/A | N/A | N/A | Available | Available | Available | N/A |
| Coverage gap | N/A | N/A | N/A | N/A | N/A | N/A | Await source | Await source |
| Debt market | N/A | N/A | N/A | Provisional | Available | Available | Available | Available |
| Equity indices | N/A | N/A | N/A | Provisional | Available | Available | Available | Available |
| Investor participation | Provisional | Provisional | Provisional | Provisional | Available | Available | Available | Await source |
| Market performance | Provisional | Provisional | Provisional | Provisional | Available | Available | Available | Available |
| Market valuation | Provisional | Provisional | Provisional | Provisional | Available | Available | Available | Await source |
| Strategy | N/A | N/A | N/A | N/A | N/A | N/A | N/A | Available |
"N/A" marks a family/year with no recorded observation at all; either the metric or programme did not yet exist for that year, or the year predates this dashboard's extraction scope. It is never used to mean "missing" for a year the family is otherwise tracked in.
Full availability detail (canonical/provisional observation counts, all 45 rows) →
Use "not applicable", not "missing", where a series (e.g. the 10X scorecard) did not yet exist.
Metric dictionary
Source register
Every chart and table on this dashboard exposes its source document, PDF page and table label.
Source hierarchy
Source hierarchy
- Latest audited financial statement or audited restated comparative: Financial statements
Canonical - Original-year detailed market table: Market, participation, indices, valuation, debt
Canonical unless internally inconsistent - Current-year detailed performance narrative/table: Current-year operational metric
Canonical when scope is explicit - Strategy scorecard: Strategy KPI only
Never substitute for audited/market metric without definition - Later five-year summary: Cross-check / provenance
Not canonical when an original-year source is available - Headline/business-model summary: Communication layer
Rejected when it conflicts with detailed source - Derived calculation: Analytical check
Publish only with formula/basis label
A hierarchy resolves precedence, not differently-defined measures.
Known limitations
- Text-layer extraction identifies possible issues; it does not verify them
- "Both sources verified" confirms transcription, not resolution
- A derived value cannot be more reliable than its least reliable input
- Hierarchy rank does not override a genuine definition conflict
Full source-hierarchy register (all 7 ranks) →
Downloads
Verified dataset
Publication-ready observations only
Complete flagged dataset
All data with status & caveats
Reconciliation Register
Full evidence record
Metric dictionary
Definitions & formulas
Source register
Document & page mapping
Change log
Corrections & status transitions
Two datasets are specified. They are not built or published in this local development stage.
- Verified dashboard dataset: publication-ready observations only (
bse_mart_download_public, 340 rows). - Complete flagged dataset: every observation with status, reconciliation ID, caveat and publication treatment (
bse_mart_download_flagged, 357 rows).
Provisional and verified data are never mixed without row-level flags.
Revision history of this release
- v1.0: approved canonical release (4 Sep 2026).
- v1.0.1: metadata correction (5 Sep 2026): seven
MISSINGrows set non-provisional; generic conflict wording removed. No numeric change. - v1.0.2: controlled augmentation (6 Sep 2026, owner-approved): three directly-reported page-40 total-market-turnover observations added (FY2025 P9.3bn, FY2024 P7.9bn, +18.1%). The 354 v1.0.1 rows are unchanged; v1.0.1 is retained as audit history.
▤ This page explains uncertainty; it is not the only place it appears. Chart-level
warnings on Pages 1-6 link back here. Canonical release bse_canonical_v1_0_2 ·
database status approved · approved_by = "Thegreat P. Malema" (2026-09-08).
Disclosure layer A2_VERIFIED. Dashboard deployment and publication are governed
separately, under GATE B (Phase 10-11).
