Zijin Mining

Refresh delta (since 2026-07-10)

The grounding upgrade. The July dossier was web-only: no filings, and every CSV held only a header. This refresh pulls the record straight from HKEXnews (stock 02899): 29 documents in filings/, covering four annual-results announcements (FY2022 to FY2025), the full 2025 Annual Report, eight quarterly or interim reports (Q1 2023 to H1 2026) and 13 material 2026 announcements. It also adds 10 quarters of CAS financials with a sourced balance sheet for the last four, the H1 2026 product split, 14 guidance rows, capexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. projects and a rebuilt model.xlsx. Where this dossier's numbers disagree with the July one, these filings win.

What changed (sourced):

1. H1 2026 was a record. Q2 was flat on Q1. H1 revenue was RMB194.18B (+15.8%), profit before tax RMB63.49B (+84%) and net profit attributable RMB39.17B (+68%). Operating cash flow was RMB55.47B (+92%). Q2 attributable profit was RMB19.09B against Q1's RMB20.08B. On a recurring basis, though, Q2 rose about 6% to ~RMB19.6B [same, p.24]. The July dossier's base case for FY2026 net profit was RMB60B. H1 alone delivered 65% of that.
2. The copper growth leg slipped. Mined copper fell 5.7% to 534kt in H1. Excluding Kamoa it rose 4.8% to 480kt. Kamoa-Kakula's 2026 plan was cut from 380–420kt to 290–330kt (100% basis). Zijin says this takes 22–57kt off its own 2026 output and "will put pressure on" the 1.2Mt target. It also admits some products are "relatively behind the scheduled progress of the annual plan".
3. US enforcement reached Zijin copper. On 16 Jun 2026 CBP issued a Withhold Release Order on copper and copper products made by Serbia Zijin Copper D.O.O. (the Bor smelter/mining company, 63%-owned, 59kt in H1). It does not name Serbia Zijin Mining (Čukaru Peki, 85kt) [interim p.29], citing six ILO forced-labour indicators. In July's bear case, sanctions spreading to copper was the "low-probability tail". It has now partly happened, at the subsidiary level.
4. The post-founder M&A record so far. On 26 Jan 2026 Zijin Gold International (the listed subsidiary) agreed to buy Allied Gold for CAD5.5B. That deal was terminated on 29 Jul 2026 because closing conditions could not be met by the outside date. In its place, a 9.2% subscription for CAD416.6M was agreed; it is still subject to TSX/NYSE approval and other closing conditions. The control deal for Chifeng Gold (~RMB18.26B) is still pending, and its long-stop date has moved from 30 Sep to 30 Nov 2026.
5. The balance sheet de-risked and payouts went up. Debt-to-assets is 49.55%, below 50% for the first time since 2012 [filings/hkex-interim-results-2026.md, p.24]. Net debt fell to RMB74.1B from RMB98.6B at end-2025. The current-portion line includes ~RMB1B of non-debt items, so a cleaner figure is ≈RMB73.0B. The payout floor for 2026–28 rose from 30% to 35%. The interim dividend is RMB0.42/share (RMB11.1B), and more than RMB20B will be paid in calendar 2026 [filings/hkex-interim-results-2026.md, p.27].
6. The Q3 date is 23 Oct 2026, not 16 Oct. The board meets on Friday 23 Oct 2026 to approve the nine-month results. Last year's Q3 report came out the same evening as the board meeting (17 Oct 2025).

What held. The structural thesis still stands. Zijin is a low-cost, fast-building, multi-metal miner with mining-business gross margins of 69% (H1 2026, up 9.1pp) [filings/hkex-interim-results-2026.md, p.34]. It is now a real three-metal story: lithium output rose 496% to 43.6kt LCE in H1. The cracks are in delivery: copper volumes, and the deal record under the new chairman. They are not in the cost position.

---

Phase A: Understand the business

Company Overview: carried, with the H1 2026 mix

Unchanged in substance; see the previous dossier. Updates: Zijin now describes itself as "focusing on gold, copper and lithium", calling gold the "monetary metal", copper the "computing power metal" and lithium the "energy metal". In H1 2026 gold was 41.2% of operating income and 41.7% of gross profit. Copper was 32.6% and 33.8%; lithium 2.2% and 3.4% [same, p.28–29]. HQ is Shanghang, Fujian. The largest holder is Minxi Xinghang State-owned Assets Investment Co. with 22.88% (A shares, 31 Dec 2025). This closes the July open item on the controlling stake. It confirms a local-SOE anchor with no other holder above 5% apart from the clearing nominees.

Supply Chain: carried

See the previous dossier. Two corrections and one addition:

• Kamoa-Kakula attributable interest is 44.19% as Zijin reports it [filings/hkex-interim-results-2026.md, copper table p.29]. That is the basis for every Zijin production figure. July's "39.6% direct" is the JV-level number.
• The asset list is longer. Manono lithium (DRC, 54.9%) started its dense-media plant in May 2026, with the smelter due by end-2026. Zangge Mining (26.2%) is consolidated. Akyem (Ghana) and Raygorodok (Kazakhstan) are contributing gold [same, p.25, p.29].
• The investing filter has changed. Management now says new overseas priorities are "neighbouring countries of China as well as other friendly countries" [same, p.45]. That is a stated retreat from Western-adjacent jurisdictions.

Competitive Advantages: carried

See the previous dossier. The cost moat shows in the numbers. H1 2026 gross margins were 79.3% on gold concentrate, 70.4% on copper concentrate and 61.8% on lithium salts [filings/hkex-interim-results-2026.md, p.34]. Unit costs are rising, though. Copper concentrate cost rose 16% YoY to RMB24,480/t, driven by lower feed grade, Middle-East-war diesel and chemical prices, longer haul distances and royalties [same, note 4, p.34].

Segments: refreshed with primary data

Zijin's reportable segments under CAS are mining, refining, trading and other. The useful cut is the product table [filings/hkex-interim-results-2026.md, p.33–34]. It is now in our figures as six-month H1 2026 rows labelled 2026-Q2.

. Prior-year margins are in brackets.

Trend and cause. Profit mix is still tilting toward gold on price: gold averaged US$4,693/oz in H1, up 53% [same, p.23]. Lithium has gone from a rounding error to a real contributor. Lithium carbonate averaged RMB163,427/t in China, up 132%, and the lithium-salt margin more than doubled. Copper is the weak leg on volume, not on price: LME copper averaged US$13,083/t, up 38.5%. Group gross margin was 37.75% (+14.0pp) and mining-enterprise gross margin 69.34% (+9.1pp) [same, p.34].

---

Phase B: Measure performance

Earnings Result: refreshed (H1 2026, printed 23 Aug 2026; Q3 due 23 Oct)

Quarterly path, RMB bn (CAS)

• Versus guidance. On 9 Jul the company pre-announced H1 attributable profit of about RMB39.1B [filings/hkex-2026-07-09-h1-profit-estimate.md]. It printed RMB39.17B, in line. No sell-side consensus for H1 was sourced, so beat or miss against consensus is n/a.
• What drove it. Price did most of the work: gold +53%, silver +139%, copper +38.5%, lithium +132% (H1 averages) [interim p.23]. Volume added gold +13.4% and lithium about 6x. Operating costs actually fell 5.5% while revenue rose 15.8%, because the low-margin refined and trading volume shrank. Refined/processed/traded gold fell 40% [interim p.28, p.35].
• Quality flags in the P&L. Investment income fell 35% "mainly due to investment losses from commodity futures contracts", partly offset by fair-value gains on futures. Hedging cost something in a rising tape. Taxes and surcharges rose 67% on resource tax. Income tax rose 134%, giving an effective rate of 21.5% versus 17.0%. Other comprehensive income was a RMB8.9B loss, largely FX translation of foreign operations as the RMB firmed [interim p.6].
• Balance sheet (30 Jun 2026). Cash was RMB91.3B, up from RMB65.6B, helped by the US$1.5B zero-coupon CB. Total borrowings plus bonds were RMB165.4B, so net debt was ~RMB74.1B, or ≈RMB73.0B excluding ~RMB1B of non-debt items in the current-portion line. That is ~US$10.9B at the 30 Jun rate (~6.79), ~US$11.05B today. Inventory was flat at RMB39.7B and trade receivables were RMB10.3B (RMB9.3B at year-end), both moving with revenue. Treasury shares rose to RMB3.46B on the A-share buyback [interim p.4].
• Cash conversion. H1 OCF of RMB55.5B was 1.42x attributable profit. H1 free cash flowFree cash flowCash left after paying to run and maintain the business. Unlike profit, it is hard to flatter with accounting choices. was RMB42.5B.
• Market reaction. The H shares were HK$29.90 at the last dossier (8 Jul). They were HK$32.16 at the latest quote, with a 52-week range of HK$26.58–46.98. The path matters. The 52-week low was intraday on 30 Jun 2026 (close HK$27.40). The stock then rallied ~30% into the print: HK$38.56 at the 21 Aug close, the last session before results. After the print it went to HK$38.44 on 24 Aug (−0.3%) and HK$37.04 on 25 Aug (−3.6%). It has since fallen ~17% to HK$32.16 as gold slid from ~US$4,700 to ~US$4,170 [price path per coordinator check against Yahoo daily closes, 2026-10-05]. A shares were CNY29.79 (30 Sep). Implied LTM P/E is ~10.8x on an H-price basis. Valuing each share class at its own price, with the A shares at an ~8.5% premium, gives ~RMB778B (~US$116B) and ~11.5x. The rest of this dossier uses the H-price basis unless stated. The market rallied into a record +68% half and then sold it as gold fell: it is pricing the gold price, not the half.

Q3 2026 print: Fri 23 Oct 2026 (board meeting; publication normally the same evening, HKT). What to watch:

1. Q3 attributable profit against Q2's RMB19.09B and Q3 2025's RMB14.57B. Gold spot was US$4,143/oz on 5 Oct, down ~6% in a month and ~26% below January's US$5,608 high. That is well under H1's US$4,693 average. Copper is higher: LME 3M was US$14,258/t on 5 Oct versus the H1 average of US$13,083. The Q3 average prices are n/a. The quarter could come in anywhere from about RMB17B to RMB20B (see Lens 11).
2. Nine-month mined copper against the 1.2Mt full-year plan. H1 was 534kt, so H2 needs ~666kt, 25% above H1's pace. Look for an explicit cut.
3. Nine-month gold against 105t. H1 was 46.7t, so H2 needs ~58t (+24% HoH), relying on the Akyem and Raygorodok ramps.
4. Lithium against 120kt LCE. H1 was 43.6kt and the Manono smelter is due only at end-2026. Track the lithium-salt margin (61.8% in H1) as lithium prices move.
5. Copper unit cost trend (+16% YoY in H1), and whether futures/hedging losses show up again in investment income.
6. Corporate events. Chifeng Gold closing (long-stop 30 Nov), any change to the Serbia WRO, Zhunuo and Manono commissioning on schedule for end-2026, and the cash position after the RMB11.1B interim dividend.

Earnings Calls / Management Communication: refreshed

There is still no transcript on the shelf. Zijin holds results briefings, including one for 2026 interim results [filings list: "Convening of 2026 Interim Results Briefing", 19 Aug 2026]. No English transcript was found. The Quartr Q1 2026 page is a summary only, with no speakers or Q&A. The read below comes from management's own disclosure text across four documents:

• Feb 2026 three-year plan. Unusually candid self-criticism. It names "the major conflict between increasing globalisation and the restrictive domestic mindset and management style", a global management system "not yet fully established", and a "shortage of globalised talents" [filings/hkex-2026-02-09-three-year-production-plan.md, p.4]. The mood is ambitious. The plan targets copper of 1.5–1.6Mt and gold of 130–140t by 2028, and a "leading global mining company" by 2035.
• Mar 2026 annual results. Triumphal on the numbers: record profit, debt falling, RMB261.9B of unused bank lines.
• Aug 2026 interim. Two moods at once. The numbers are celebratory ("economic indicators reached new record highs"; Forbes top-3 metal miner). The operating language is defensive. Management admits being "behind the scheduled progress of the annual plan", says it had to make up shortfalls caused by Kamoa and by "regional mining work safety inspection campaigns" in China, and credits "strategic price locking" with keeping the overseas supply chain running through the Middle East war [interim p.24–26].
• New vocabulary. Copper as the "computing power metal". "Increasing production" written into the overarching principle. A stated "neighbouring countries... friendly countries" investment filter. "Market value management" (a buyback and dividends agenda).
• Stopped saying. The founder-era "unprecedented risk" framing of Aug 2025 has become a standard risk section: resource nationalism, rising mining taxes, metal prices, ESG compliance [interim p.45].

Tone trend: expansion confidence held, and delivery confidence slipped. This management team now admits execution slippage in writing, which the July dossier had not seen.

Comps: carried, Zijin row updated

The peer table in the previous dossier is carried unchanged. No peer multiple was re-sourced this run, so treat them as July-dated. Zijin's own row:

model.xlsx → Comps sheet has the scaffold. Peer multiples need re-sourcing before the next pass.

Stock-Price Catalysts: refreshed

The July list is carried; see the previous dossier. New since then:

• Jan 2026 peak. Market capMarket capitalisationThe share price multiplied by the number of shares. What the market says the equity is worth. topped RMB1.1T [filings/hkex-2026-02-09-three-year-production-plan.md, p.2]. The H shares closed at HK$46.14 on 29 Jan, when the CB priced at HK$63.30, a 37% premium [filings/hkex-2026-01-30-convertible-bonds.md]. The 52-week high was HK$46.98. Gold peaked at US$5,608 the same month.
• Feb–Jun 2026 drawdown to the HK$26.58 52-week low (intraday 30 Jun), as gold fell from above US$5,500 to below US$4,000 at one point (the company's own description, interim p.23). The two moved together, which again shows this stock trades on the gold price.
• 16 Jun 2026: CBP WRO on Serbia Zijin Copper. No dated >5% move is sourced for that day: n/a.
• 29 Jul 2026: Allied Gold deal terminated. No termination fee either way [filings/hkex-2026-07-29-allied-gold-termination.md].
• 30 Jun → 21 Aug 2026: ~+30% rally into the print, from the intraday 52-week low of HK$26.58 (close HK$27.40) to HK$38.56.
• 23 Aug 2026: H1 results +68%. The stock did not rise on the print: HK$38.56 (21 Aug) → HK$38.44 (24 Aug, −0.3%) → HK$37.04 (25 Aug, −3.6%). It then fell ~17% to HK$32.16 by early October as gold slid from ~US$4,700 to ~US$4,170 [price path per coordinator check against Yahoo daily closes]. A sell-the-news move followed by a gold-led drawdown.
• Next: Q3 print on 23 Oct 2026. See Lens 5 for what to watch. Then Chifeng closing by 30 Nov, and Zhunuo and Manono commissioning by end-2026.

Pattern (unchanged and stronger): the stock follows gold first, copper second and China/sanctions sentiment third. Company-specific beats barely move it.

---

Phase C: Judge people & books

Management: refreshed

• The new team is operating. Zou Laichang is Chairman, Lin Hongfu is President, and 2026 is the ninth Board's first year. Governance was restructured: an Audit and Supervision Committee replaces the old supervisory committee, and a "joint chief officer" role now covers internal supervision [interim p.26].
• Capital-allocation record, year one. (a) Allied Gold: a CAD5.5B (~US$4.0B) deal at CAD44/share was terminated because conditions could not be met, Zijin Gold International then agreed a 9.2% subscription at CAD32.55, about 26% below the deal price; it still needs TSX/NYSE approval [filings/hkex-2026-07-29-allied-gold-termination.md]. The filing does not say which conditions failed. Discipline or failure? Unproven either way. If it completes, the fallback stake comes at a lower price than the original offer. (b) Chifeng Gold: RMB18.26B for ~25.85% and control, at a 1.3% premium on the A shares and 83% of the 60-day average on the new H shares [filings/hkex-2026-03-23-chifeng-gold-acquisition.md]. That is a disciplined price, but the deal has not closed (long-stop now 30 Nov). (c) Smaller bolt-ons: Gansu Northwest Gold (72%, ~RMB2.4B) and Lianrui tungsten-tin ("approximately RMB900 million", partly through a judicial auction) [interim p.25].
• Capital returns rose. The payout floor went from 30% to 35% for 2026–28. The interim dividend was brought forward. The A-share buyback lifted treasury shares to RMB3.46B [interim p.4, p.27].
• Incentives. The 2026 ESOP was subscribed by executives "actual, undiscounted", with part of future incentive pay given up. Bonuses for executives and core staff are now deferred and vest against future share-price performance [interim p.27]. That is better alignment than the July dossier could verify.
• New red flag (minor, watch). On 15 Sep 2026 a wholly-owned subsidiary committed RMB1B (20.4%) to Xiamen Shidai Shenyuan venture fund, which targets AI and embodied AI. Kunkun Investment, 100% owned by founder Chen Jinghe and his spouse, is a co-LP with RMB200M [filings/hkex-2026-09-17-connected-transaction-fund.md]. It is small. Still, it is off-core, and it is a related-party co-investment with the departed founder, whom the filing calls the "former chairman" (the "honorary chairman" title comes from a web source: Zijin blog-detail-122530, cited in the July dossier).

Forensic Red Flags: refreshed

• Earnings quality is good. OCF/attributable profit was 1.42x in H1 2026 and 1.46x in FY2025 (RMB75.4B / 51.8B) [interim p.7; annual-results-2025 p.82]. Receivables and inventory are not outrunning revenue.
• Watch items. (1) Q4 2025 capex jumped: RMB15.8B against ~RMB5B in each earlier quarter. Whether that is year-end catch-up or a reclassification is unexplained. (2) The derivatives book. Board-approved speculative trading limits are small (RMB300M + US$100M margin), but group hedging produced an investment-income loss in H1 [interim p.35, p.41]. (3) Disclosure hygiene. On 6 Sep 2026 the company published a correction for "typographical errors in the notes to the financial statements" [HKEX listing, 2026-09-06; document not pulled]. (4) Bargain-purchase gains booked in non-operating income on 2026 acquisitions [interim p.36]. (5) Total equity of RMB273.1B includes RMB67.2B of minority interests, so group ratios flatter the parent.
• Provisions. Litigation provisions are only RMB33M (Serbia Zijin Copper, Guyana Goldfields, Longking, La Arena). Environmental rehabilitation provisions rose to RMB8.58B from RMB4.88B, mostly through acquisitions. The only quantified pending case is RMB54.6M (Longking) [AR 2025, p.383].

Regulatory findings (required sub-section):

• SEC LR/AAER: none possible, as Zijin has no CIK.
• US trade enforcement: MATERIAL and escalating. Zijin has been on the UFLPA Entity List since January 2025. New: a CBP Withhold Release Order, 16 Jun 2026, on copper and copper products from Serbia Zijin Copper D.O.O. It cites abuse of vulnerability, withheld wages, intimidation, restricted movement, retained identity documents and excessive overtime. It is the fourth WRO of FY2026. Serbia Zijin Copper produced 59kt of electrolytic copper in H1 2026 [interim p.29]. Its direct US sales exposure is n/a. The order names Serbia Zijin Copper D.O.O. (Bor, 63%-owned) only, not Serbia Zijin Mining (Čukaru Peki, 100%, 85kt in H1). The importance is precedent: it is the first US action on Zijin copper, and on a non-Chinese subsidiary.
• Serbia social licence. Land-expropriation and pollution allegations (Krivelj and Ostrelj villages) continue.
• Item 3 equivalent (AR 2025): the pending-litigation note lists one case (above). Nothing material.
• Summary: no sign of accounting fraud. The forensic risk is still governance, ESG and geopolitics, and the geopolitical part has moved from listing to enforcement.

---

Phase D: Project & stress-test

Forward Projection: refreshed

The model. model.xlsx was rebuilt on 2026-10-05 from 10 quarters (2024-Q1 to 2026-Q2). Its opening balance sheet is sourced: yes (30 Jun 2026 filing values). our model reports no computed values: the workbook has never been opened, and the share price is blank. Three seeding caveats before anyone trusts it:

• (a) The CAS "operating margin" includes investment and fair-value income, so the opex driver clamps to 1%.
• (b) Total equity and the model's earnings line are group-level, including minorities of ~21% of net profit.
• (c) D&A is an: FY2025 D&A ÷ 4.

No our model entry was logged (unattended).

Arithmetic projection (net profit attributable, RMB):

• Starting point: H1 2026 RMB39.17B. Attributable profit per RMB of pre-tax profit ≈ 0.62. H1 mined-gold sales were RMB44.47B and mined-copper sales RMB38.84B [interim p.33].
• FY2026 base ≈ RMB77B (+49%), EPS ≈ RMB2.90. H2 gold average ~US$4,300 (−8% vs H1) costs ~RMB3.7B pre-tax per half (−RMB2.3B attributable). H2 copper ~+5% adds ~RMB1.9B pre-tax (+RMB1.1B attributable). Volumes are roughly flat HoH. So H2 ≈ RMB38B and FY ≈ 39.2 + 38. Shares are 26.59B; CB dilutionDilutionIssuing new shares, so each existing share owns a smaller slice of the same company. of ~185M shares is about 0.7%.
• FY2026 bull ≈ RMB81B, EPS ≈ RMB3.05. Gold back to an H1-like ~US$4,700, copper at ~US$14k, lithium salts staying strong as Manono adds.
• FY2026 bear ≈ RMB72B, EPS ≈ RMB2.71. Gold at ~US$3,900 in H2 (−17%: −RMB7.6B pre-tax, −RMB4.7B attributable), plus a further Kamoa or safety-stoppage hit of about −RMB1B.
• FY2027–28 (base) ≈ RMB85B and ≈ RMB95B. This assumes flat metal prices at the base FY2026 deck with volume growth toward the 2028 plan: Kamoa back to 380–420kt in 2027, Julong at 300–350kt, Zhunuo 76kt and Manono on stream. Haircut it if the copper plan keeps slipping.
• Valuation read. At HK$32.16 (RMB27.46 at 0.8539) the stock is ~9.5x base FY2026e, 9.0x bull and 10.1x bear on the H-price basis. Consensus targets: the H-share average is HK$51.37; the A-share average is CNY47.99 (15 analysts, all Buy, range CNY39–58). Both imply ~60% upside. Consensus earnings estimates are n/a. The only one seen was First Shanghai's 2026 net profit of RMB69.8B (Oct 2025, stale). July's FY2026 base of RMB60B was far too low: H1 alone ran at a ~RMB78B annual rate.

Bull vs Bear: refreshed

Bull. Zijin earned RMB67.7B attributable in the last four quarters and trades at ~11x that, ~10x 2026e and ~7x EV/EBITDA. Mining gross margins are 69%. Net debt is RMB74B, against RMB55B of OCF per half. Payouts are rising to ≥35%, with buybacks on top. The franchise now has three legs. Gold is still the world's #1 by Forbes ranking. Copper is set for 2027 recovery as Kamoa returns to 380–420kt, with Julong and Zhunuo adding. Lithium went from 7kt to 44kt in a half at a 62% salt margin. The price already assumes gold stays near US$4,100. If gold merely holds and copper (the "computing power metal") stays above US$13k, 2026 profit lands near RMB75–80B and 2027 grows on volume.

Bear. This is a price-taker whose 2026 profit is roughly two-thirds metal price. Gold is ~26% off its January high and still falling month on month, and every US$100/oz is ~RMB1B a half pre-tax. The volume story, the part that was meant to be idiosyncratic, is slipping: copper is down 5.7% YoY in H1, the 1.2Mt target is under pressure and the company concedes it is behind plan. Unit costs are rising (copper concentrate +16%). Political risk moved from listing to enforcement with the Serbia WRO. The new chairman's first large deal failed, the next one has been delayed, and the founder is co-investing with the company in an off-core AI fund.

Pre-mortem (it is April 2028 and the thesis broke). Gold kept falling toward US$3,500 as real rates rose. Kamoa's recovery slipped again. Copper never reached 1.5Mt by 2028 and the plan was quietly re-based. CBP extended WRO-style action to other subsidiaries, and Western banks pulled back from Zijin-led deals. Chifeng closed at the top of the gold cycle. Profit fell back toward RMB50B and the multiple did not expand: a round-trip to the 2025 starting point.

Are multiples too high? No. At ~10x a record year, the risk is the E, not the multiple.

Contrarian view. The market treats the January–October drawdown as "gold rolled over, sell the gold miner". It is missing that Zijin's mix is drifting away from gold: copper ex-Kamoa is still growing, lithium has gone 6x and moly/tungsten are rising. And the balance sheet now lets it buy counter-cyclically, as it did in 2018–20, if the new team keeps the discipline the Allied termination may or may not show.

Devil's Advocate: refreshed

• The volume story is a plan, not a record. Every 2026 headline metric is behind run-rate at the half: copper 534kt against 1.2Mt, gold 46.7t against 105t, lithium 43.6kt against 120kt. Hitting all three needs an H2 that is 24–75% bigger than H1. The company's language ("behind scheduled progress") is the first written admission. Expect cuts on 23 Oct or with FY results.
• Concentration risk proved itself twice. Kamoa (44.19%, 54kt attributable in H1) failed again. Julong, now the biggest copper mine (134kt in H1), is on the Tibetan plateau and exposed to the "regional mining work safety inspection campaigns" that already stopped some domestic mines in H1.
• The political discount is now enforcement, not sentiment. A WRO on copper from the Bor subsidiary (Serbia Zijin Copper) creates a template: a subsidiary outside China, a product (copper) that is not a "Xinjiang good", and evidence gathered by NGOs. If Buriticá (Colombia) or the DRC assets draw similar scrutiny, the offtake and finance channels narrow further.
• Capital allocation without the founder. One large deal failed and one is delayed. A related-party AI-fund co-investment with the founder sits alongside. The plan calls for more "controlling stakes in listed companies". That is the classic pattern of a mining-cycle top, and the gold cycle may have topped in January.
• What has to hold for HK$32. Gold ≳ US$3,800, copper ≳ US$11k, no further WRO spread, and 2026 copper at least ~1.1Mt. If earnings disappoint 25% (FY2026 ~RMB58B, i.e. gold at ~US$3,500 in H2 plus volume misses), the stock is at ~12.6x on a falling trend. A de-rating to 9x would mean ~HK$23.0. That is about −29%.
• The single permanent-impairment scenario is unchanged: G7-wide import or finance exclusion. Its probability rose this year (listing in Jan 2025, then WRO in Jun 2026).

Management Questions: carried, re-ordered

See the previous dossier for the original 15. For the 23 Oct print, the first five by information value are:

1. Will you re-base the 2026 targets (1.2Mt copper, 105t gold, 120kt LCE)? If not, which assets close an H2 gap 25–75% larger than H1?
2. The Serbia WRO: what share of Serbia Zijin Copper output reaches US-linked buyers? What remediation is under way? Have any banks or offtakers changed terms?
3. Why did the Allied Gold conditions fail? What is the walk-away rule for Chifeng if the 30 Nov long-stop slips again?
4. Copper unit cost rose 16% YoY. How much of that is grade, which reverses, and how much is royalty and energy, which does not?
5. What is the governance rationale for co-investing with the former chairman's private vehicle in an AI venture fund?

---