Frontierby eninvs

Language: EN · RU

← Company

KMGZ 2025-12-31 FY — report review

Status: OK; Currency: KZT; Amounts unit: millions; Forms:

Report published: Not stored for this period — set financial_report_date on the row (EDGAR filingDate, KASE change_date, or manual_catalog).

Full financial report: Report (PDF)

PDF (local): data/raw_pdfs/KMG/2025-12-31_FY_12m2025-conso-fs-kmg-eng.pdf

To recalculate statement detection and previews from the PDF, use this link The default link runs in the background: a status panel shows phase, elapsed time, rough ETA, CUDA vs CPU, and OOM hints, then loads the finished report. Heavy mode with refresh does this automatically so reverse proxies do not return 502. Add &sync=1 only for one long blocking request (not recommended). You can use ?refresh=1, ?recalc=1, ?nocache=1, or ?recompute=1 on the URL. (block in the browser until done: synchronous refresh)

Default view is fast (metric table + statement page numbers only) so reverse proxies do not time out. Use ?heavy=1 when you need embedded page images and Camelot tables.

Metric mapping (value → extracted evidence)

Metric values use dashboard units where applicable; evidence is the stored snippet from the PDF text layer or OCR used during extraction.

MetricValueEvidence / page extract
Revenue9 371 488Row: revenue (mln KZT, batch apply) · dashboard=9,371,488.000 mln — [DeepSeek] revenue (mln KZT, batch apply)
Operating profit778 213Row: operating_profit (mln KZT, batch apply) · dashboard=778,213.000 mln — [DeepSeek] operating_profit (mln KZT, batch apply)
D&A723 977Row: da (mln KZT, batch apply) · dashboard=723,977.000 mln — [DeepSeek] da (mln KZT, batch apply)
EBITDA1 502 190Row: ebitda (mln KZT, batch apply) · dashboard=1,502,190.000 mln — [DeepSeek] ebitda (mln KZT, batch apply)
Net profit1 072 000Row: KMG FY2025 IR/KASE press release: net profit incl. JV share KZT 1,072bn (vs 1,094bn FY2024); filled missing latest np · dashboard=1,072,000.000 mln — KMG FY2025 IR/KASE press release: net profit incl. JV share KZT 1,072bn (vs 1,094bn FY2024); filled missing latest np
Cash1 198 185Row: cash (mln KZT, batch apply) · dashboard=1,198,185.000 mln — [DeepSeek] cash (mln KZT, batch apply)
Debt short295 100Row: debt_short (mln KZT, batch apply) · dashboard=295,100.000 mln — [DeepSeek] debt_short (mln KZT, batch apply)
Debt long3 355 819Row: debt_long (mln KZT, batch apply) · dashboard=3,355,819.000 mln — [DeepSeek] debt_long (mln KZT, batch apply)
Net debt2 534 827Components: short debt 295 100 + long debt 3 355 819 + other financial liab. 0 + NCI 82 093 − cash 1 198 185 = net debt 2 534 827.Row: net_debt (mln KZT, batch apply) · dashboard=2,534,827.000 mln — [DeepSeek] net_debt (mln KZT, batch apply)
Operating CF1 900 424Row: operating_cash_flow (mln KZT, batch apply) · dashboard=1,900,424.000 mln — [DeepSeek] operating_cash_flow (mln KZT, batch apply)
Investing CF-1 175 772Row: investing_cash_flow (mln KZT, batch apply) · dashboard=-1,175,772.000 mln — [DeepSeek] investing_cash_flow (mln KZT, batch apply)
Assets18 848 445Row: total_assets (mln KZT, batch apply) · dashboard=18,848,445.000 mln — [DeepSeek] total_assets (mln KZT, batch apply)
Equity12 386 325Row: total_equity (mln KZT, batch apply) · dashboard=12,386,325.000 mln — [DeepSeek] total_equity (mln KZT, batch apply)

Consistency checks · All checks passed

Balance sheet identity (A = L + E)TA (18,848,445) ≈ TL (6,462,120) + TE (12,386,325); residual +0 within 1%.
Net debt formulanet_debt 2,534,827 matches |debt_short|+|debt_long|+|other|+|NCI|−|cash| = 2,534,827.
EBITDA = OP + D&AEBITDA (1,502,190) ≈ OP (778,213) + D&A (723,977) = 1,502,190.
Net profit vs operating profitNet profit (1,072,000) sits within a plausible band vs operating profit (778,213).
Cash ≤ total assetsCash (1,198,185) ≤ total assets (18,848,445).

Statement pages (discovery)

FormPages
P&L12, 13
BS14, 15
CF16, 17

Statement previews & reconstructed tables

Embedded PDF thumbnails and Camelot reconstruction are omitted in the default view to avoid gateway timeouts and huge HTML. Open full previews & tables (?heavy=1).

Formulas used