§ سجل التغييرات
إصدارات المنهجية
كل تغيير في طريقة تحليلنا للأسهم، مؤرَّخاً ومُفسَّراً. لا نعيد كتابة المنهجية بصمت أبداً.
- v3.2
- Versioned the methodology itself, so every report records the exact methodology it was produced under
- Made accounting-quality checks source-aware before earnings are trusted
- Extended long-horizon financial history and regulatory-filing coverage behind each analysis
- Added per-ticker input-coverage scoring so thin evidence is caught before an analysis begins
- Blocked silent substitution on valuation-critical fields
- Added third-party fair-value cross-checks alongside the house range
- v2.5
- Updated the public methodology version for the May 2026 research cycle
- Improved context-aware calibration across business maturity, industry structure, financial strength, and evidence quality
- Strengthened public-output hygiene so private calibration labels and third-party benchmark names stay out of reader-facing reports
- Added stricter release checks for valuation coherence, data provenance, translated labels, and audit-appendix readability
- v2.4
- Added stronger business-archetype treatment for valuation context
- Tightened accounting-quality review before earnings are trusted
- Expanded sensitivity analysis beyond single-variable checks
- Improved fair-value ranges so model convergence is clearer to readers
- v2.3
- Capital-intensive archetype split out from cyclical
- Reverse-DCF added as a primary lens for hyper-growth names
- Earnings decision tree formalized as a standard section
- v2.2
- Peer-comparison set expanded to 6 names per ticker
- Added 4 new locales (Vietnamese, Thai, Persian, Turkish)
- Composite weights rebalanced for financial archetype
- v2.1
- Initial public release
- Coverage of 800 tickers
- 14 supported locales