Jiangxi Copper

Phase A — Understand the business

Company Overview

Jiangxi Copper Company Limited is China's largest integrated copper producer and the largest copper-cathode smelter in the world by single-site scale (the Guixi smelter). The business is a fully vertically integrated copper chain — exploration → mining → ore dressing → smelting → refining → fabrication (copper rod/wire) — plus recovery of the precious and rare metals that ride along in copper concentrate (gold, silver, molybdenum, selenium, tellurium, rhenium, platinum-group).

How it actually makes money — three engines:

1. The processing engine — toll-convert purchased copper concentrate into refined cathode, historically earning a treatment & refining charge (TC/RC). This is the bulk of revenue but, at 2026 benchmark TC/RC of $0/t, earns approximately zero gross profit. The business's nominal core is presently a non-earning conversion utility.
2. The by-product engine — gold, silver and sulphuric acid extracted from the same concentrate. Historically a "rounding error" supplement; at gold ~$4,600/oz and sulphuric acid ~$137/t it has become the margin of survival.
3. The resource engine — owned mines (the Dexing open-pit being the crown jewel), the Humon gold-processing subsidiary, and minority stakes in offshore miners (First Quantum, SolGold 12.19%). Provides self-supplied concentrate and optionality on the copper price.

Headline scale (2025): revenue RMB 542.7B (~$76B), refined cathode-copper output ~2.37Mt target (2024 actual 2.29Mt), gold output ~118–139t, silver ~1,200t, sulphuric acid 6.5Mt. It is roughly 77% of the output of China's listed copper smelters — a national champion, not a marginal player.

Customer/contract structure. Output is bulk industrial commodity sold into China's cable, construction, power-grid, EVEnterprise valueWhat it would cost to buy the whole company: its market value plus its debt, minus the cash you would get with it. Often very different from market cap. and electronics supply chains. There is no customer concentration in the SaaS sense — copper cathode is fungible and priced off SHFE/LME. The binding contract variable is not a customer but the annual concentrate-supply benchmark (TC/RC) Jiangxi negotiates with global miners (Antofagasta et al.), which sets the toll it earns and historically templates the whole Chinese industry. our figures is empty (web-only); concentration risk lives on the input side (concentrate supply), not the output side.

Supply Chain

Map, upstream → company → downstream, with named stakeholders:

Upstream (copper concentrate — the binding input):

• Self-mined concentrate: Dexing open-pit (Jiangxi province) is the anchor asset; H1-2025 mine output was ~99,300t of contained copper in self-produced concentrate, implying a full-year self-mined run-rate around ~200,000t vs ~2.37Mt of cathode — i.e. self-sufficiency on the order of ~8–10%. This is the single most important number in the entire dossier: ~90% of feed is purchased concentrate, so Jiangxi is overwhelmingly a toller, not a miner.
• Purchased concentrate: sourced globally — Chilean (Antofagasta, Codelco), Peruvian, African (First Quantum's Zambian/DRC output, in which Jiangxi holds a strategic stake), and the global concentrate spot market. The 2026 benchmark TC/RC was struck at $0/t with Antofagasta (down from $21.25 in 2025), and spot TCs went negative — minus ~$42/t in Nov 2025.

Midstream (the company): Guixi smelter (world's largest single-site, advanced automation) + other smelting/refining + Humon (gold) + fabrication (copper rod).

Downstream (buyers): China's State Grid / power utilities, cable & wire makers, construction, EV/auto, electronics, and — increasingly the demand narrative — AI data-centre power and cooling infrastructure. Gold by-product flows to the PBoC/jewellery/investment channel; sulphuric acid to fertiliser and industrial chemicals.

Chokepoints & single-source dependencies:

• The concentrate chokepoint is global mine supply. With ~90% of feed purchased and Chinese smelting capacity (16–17Mt) having badly outrun global mine growth, Jiangxi competes with every other Chinese smelter for a scarce input — which is why TC/RC collapsed. This is a structural, not cyclical, squeeze on the processing engine.
• By-product feedstock is itself a function of concentrate access — the gold/silver "save" depends on continuing to win by-product-rich concentrate at scale.
• Geographic concentration of owned resources is domestic (Dexing); offshore resource stakes (First Quantum, SolGold, the stalled Aynak JV in Afghanistan) are minority and largely non-producing.

This lens is names-or-it-didn't-happen and the names are here: Antofagasta, Codelco, First Quantum, SolGold, MCC (Aynak partner), State Grid. The chain's fragility is the input side.

Competitive Advantages (moats)

Real, durable moats:

• Scale & cost position in smelting. Guixi's world-leading single-site scale and automation give Jiangxi among the lowest conversion costs in the industry — which is exactly why it can survive zero-TC/RC economics that would shutter sub-scale smelters. The 2026 industry plan for Chinese smelters to cut output >10% hits marginal capacity first; Jiangxi is structurally a survivor.
• By-product-rich feed + recovery capability. The Humon gold operation and decades of metallurgical know-how let Jiangxi monetise the precious-metal and acid credits that turn a zero-margin toll into a profit. Competitors without by-product access cannot.
• State backing & benchmark-setting status. As a ~45.8% Jiangxi-SASAC-controlled SOE and the customary first-mover in annual TC/RC talks, it has privileged access to credit, concentrate, and policy support. When Beijing "firmly opposes negative treatment charges", Jiangxi is on the right side of that intervention.
• Vertical integration. Mine-to-rod presence captures margin at multiple stages and provides a (small but real) self-supplied concentrate cushion.

Where the moat is weak (critical):

• Bargaining power has inverted. The whole point of the TC/RC collapse is that miners now have pricing power over smelters, not the other way round. Jiangxi's "moat" as a toller is precisely the thing being eroded — its scale lets it endure the squeeze, not escape it. A scale advantage in a structurally loss-making activity is a defensive moat, not an offensive one.
• Commodity output, zero pricing power downstream. Cathode is a price-taker product.
• The bull's reframing (Lens 12) is that the true moat is no longer "smelting scale" but "access to gold and acid at China-champion scale" — a far better business if you believe by-product prices hold.

Segments

our figures is empty (web-only) — every figure below is ``, drawn from the 2024/2025 annual disclosures.

Trend & cause. Revenue is flat-to-slightly-up and overwhelmingly copper by the top line — but the top line is the least informative number here, because cathode revenue is mostly pass-through metal value at near-zero conversion margin. The economically meaningful split is by gross profit, not revenue, and that is where the gold/acid 14%-of-revenue segment is doing a hugely disproportionate share of the profit lifting as TC/RC went to zero. Group gross margin was ~3.4% in 2024 — razor-thin, and arithmetically impossible to sustain on copper conversion alone at $0 TC/RC, which tells you the by-product engine is carrying it. The deceleration to watch is not revenue growth (irrelevant at this margin) but gross margin direction: >3.2% = by-products working; <2.8% = the offset is fading. By-product operating-income detail is n/a (not segment-disclosed at that granularity in English filings).

---

Phase B — Measure performance

Earnings Result (FY2025, reported Apr 2026)

The latest annual print (Jiangxi reports semi-annually; FY2025 results filed Apr 2026):

(2024 revenue is variously reported as RMB 514.9B and RMB 520.9B across sources — a translation/restatement discrepancy I flag rather than resolve; both round to "~RMB 515–521B".)

What drove it. PBT (+15.3%) grew faster than attributable net profit (+6.8%) — consistent with tax/minority drag and the RMB 1.53B impairment haircut. The earnings quality story is that profit rose despite the processing engine collapsing to ~zero margin, which can only be explained by record gold and sulphuric-acid by-product prices offsetting the TC/RC wipeout. That is the entire thesis in one data point: copper conversion died, earnings still grew.

Margins. Group gross margin ~3.4% (2024); net margin ~1.3–1.4%. These are structurally tiny — Jiangxi is a high-revenue, low-margin throughput machine, so EPS is extremely operationally levered to small moves in by-product prices and TC/RC.

Balance-sheet flags. Debt-to-asset crept to 57% — not alarming for an SOE, but rising. Management noted operating cash flow fell because higher metal prices inflated inventory cost — a working-capital drag that is benign (price-driven, not collections-driven) but worth tracking: at these revenue levels, inventory swings are large in absolute RMB.

Market reaction. Not a single-print event — the stock is up ~200% over the trailing year on the cumulative gold + copper-price re-rating, then sold off ~42% from its 53.75 HKD high to 30.92 HKD by 2026-06-30. The tape says the market priced in the by-product windfall, then began doubting its durability.

Earnings Calls (sentiment trend)

transcripts/ is empty and Jiangxi (HK-listed SOE) does not hold US-style quarterly earnings calls with public Q&A transcripts; investor communication is the semi-annual results announcement + AGM. So a multi-call sentiment delta in the Western sense is n/a. What can be read from the management commentary trend:

• Recurring themes: resource security (offshore stakes — SolGold raised to 12.19% in Mar 2025), cost control / "lean smelting," and ESG/automation at Guixi.
• A notable signal of stress: the head of Jiangxi Copper's international trading arm (Su Li) stepped down — a personnel change in the unit most exposed to concentrate-procurement and hedging, in the exact year TC/RC went negative. Read it as the trading/procurement function being under pressure.
• What they've stopped emphasising: TC/RC as a profit driver — the narrative has quietly shifted from "smelting margins" to "integrated resource + by-product value," which mirrors the bull reframing.

Comps

Peer table — global integrated copper / China copper-gold names. Multiples are `` with source/date or n/a. None fabricated. Jiangxi's own line uses the H-share (0358).

Read. Jiangxi (H) at ~11.8x P/E and 4.4% yield is the cheapest line in the table on a P/E basis, in line with Zijin (~12.3x) and far below the Western miners (FCX ~23x fwd, SCCO/Antofagasta high-teens-plus EV/EBITDA). Two structural reasons for the discount, both real: (1) it is a toller, not a miner — lower-quality, lower-margin earnings than FCX/SCCO; (2) H-share/SOE/China governance discount — the H-share trades at roughly half the A-share price, a persistent dual-listing dislocation. The bull case is that the gap to Zijin is unjustified given Jiangxi's by-product leverage; the bear case is that toller economics deserve a discount to real miners. Jiangxi's own EV/EBITDA is not cleanly sourced and is left n/a rather than guessed.

Stock-Price Catalysts (last ~5y, >5% moves)

Mostly ``; the pattern matters more than each date.

• 2025 melt-up (+~200% YoY): driven by the dual tailwind of record gold ($4,600/oz) and copper near record highs ($10k–$13k/t) colliding with a market that still modelled Jiangxi as a "troubled smelter awaiting TC/RC recovery" — a classic re-rating as identity caught up to earnings.
• +13.3% single-day (29 Aug 2025) on the ADR/foreign line — an example of the leverage to commodity-price/news prints.
• TC/RC headlines ($0 benchmark Nov 2025; negative spot) — counter-intuitively, the stock rose through the worst smelting news of the decade, because the gold/acid offset dominated. The single most revealing fact about this name: the market reacts to gold and sulphuric-acid prices far more than to copper-conversion margins.
• ~–42% from the 53.75 high into mid-2026 — a gold-price wobble / profit-taking drawdown, consistent with the thesis that the stock is now effectively a levered gold + acid call option wearing a copper logo.

Pattern conclusion: this is no longer trading as a smelter. It trades as a China-listed, leveraged precious-metals + chemicals proxy. Underwrite the by-product prices, not the TC/RC.

---

Phase C — Judge people & books

Management

• Who: Executive Chairman Zheng Gaoqing; GM & Deputy Chairman Zhou Shaobing. Controlled ~45.83% by Jiangxi Copper Corporation, ultimately Jiangxi Provincial SASAC. This is a state-owned enterprise run by appointed managers, not founder-operators.
• Track record: institutional rather than personal — Jiangxi has been built into the world's largest single-site smelter and China's #1 copper producer over decades. The relevant managerial judgement call of the moment — leaning into resource security (offshore stakes) and lean smelting while TC/RC collapses — looks directionally correct.
• Skin in the game: effectively nil personal ownership; incentives are state/CCP-administrative, not equity. This is the central governance caveat: management answers to Jiangxi SASAC and national policy goals (copper supply security, employment, provincial GDP) before it answers to minority H-shareholders. Capital may be deployed for strategic-national reasons that are NPV-negative for outside holders.
• Capital-allocation history — mixed-to-poor on offshore M&A:
• Mes Aynak (Afghanistan), 25% stake — signed 2007/08, still not producing in 2026, ~$2.83B headline JV with MCC, stalled for ~17 years on security/contract. A textbook strategic-but-uneconomic SOE bet.
• SolGold 12.19% (raised Mar 2025, ~$18m) and First Quantum strategic stake — concentrate-security plays, sensible in direction but minority and illiquid.
• RMB 1,530m impairment in 2025 — confirms the offshore/asset book carries write-down risk.
• Dividend: RMB 1.00 DPS, ~4.4% yield on the H-share — a genuine, meaningful cash return, which is a point in management's favour and atypical of a value-destroying SOE.
• Red flags: (1) the historic Chinese-copper-complex association with trading/derivatives blow-ups (the broader sector's State Reserves Bureau scandal; a Mitsubishi-counterparty copper fraud in China) — not Jiangxi-specific, but the international trading arm head departed in 2025, in the unit most exposed; (2) related-party dynamics with the SASAC parent (intercompany concentrate/financing) are inherent to the structure; (3) Dexing environmental litigation/pollution history.
• Archetype: professional state-appointed managers of a strategic national asset — implies stability, cheap funding, policy tailwind on supply security, but a permanent agency gap between national objectives and minority-shareholder returns.

Forensic Red Flags

our figures empty → balance-sheet forensics are coarse (web-only), but the structural risk vectors are clear:

• Margin fragility / earnings quality. A ~3.4% gross margin means tiny absolute swings in metal prices or TC/RC flip the P&L. Reported profit is disproportionately by-product-derived at present — high-quality cash (gold/acid sell readily) but low-durability if commodity prices mean-revert. The "earnings beat despite smelting collapse" is real but is a commodity-price artefact, not an operational improvement.
• Inventory / working capital. Management explicitly flagged operating cash flow down on inventory cost inflation. At RMB 543B revenue, metal inventory is enormous; rising prices inflate carrying value and can mask weak underlying cash conversion. Watch OCF vs net income — a widening gap would be the first forensic warning.
• Impairments. RMB 1.53B asset-impairment provision in 2025, with stalled offshore assets (Aynak) a standing write-down candidate. Goodwill/long-lived-asset risk on the international book is real.
• Receivables / off-take: n/a at line-item granularity.
• SBC / non-GAAP gaming: essentially not applicable — Chinese SOE, no material stock-comp culture; the usual US tech red flag is absent here.
• Disclosure quality: semi-annual, IFRS-style HKEx filings, but English-language granularity is limited and there is no SEC oversight layer. Treat segment/by-product detail as management-reported and unaudited-by-Western-standards.

Regulatory findings (required sub-section).

• SEC enforcement (EDGAR LR + AAER): None possible — Jiangxi Copper has no CIK; it is not an SEC registrant, so EDGAR EFTS returns zero by construction (per regulatory/regulatory-findings.md, generated 2026-06-30). This is an absence of jurisdiction, not an absence of risk.
• Non-SEC enforcement (web search): No material FTC/DOJ/FDA/CFPB-equivalent Western enforcement action found against Jiangxi Copper specifically. The relevant exposures are (a) Chinese environmental enforcement / litigation around the Dexing mine (pollution case studies, NGO reports 2015–), and (b) sector-level trading/fraud history in the Chinese copper complex (State Reserves Bureau scandal; 2024 Mitsubishi copper-fraud loss in China) — context, not direct findings against Jiangxi.
• Item 3 (Legal Proceedings): n/a — no 10-K exists (foreign filer, no EDGAR). The HKEx annual report carries the equivalent contingencies/litigation note; not separately extracted here (web-only), flagged for a future filings pull if the company is ever brought onto the research shelf.
• Net: No material Western-regulatory findings — verified via SEC EDGAR EFTS (zero by no-CIK), web search, and the regulatory-findings file as of 2026-06-30. The genuine governance/legal risk is domestic (environmental, SOE related-party, no minority-protection backstop) and is captured in Lens 9/13, not in any enforcement database.

---

Phase D — Project & stress-test

Forward Projection (FY2026E–FY2028E EPS)

Built bottom-up from FY2025 actuals (EPS RMB 2.14) + the structural drivers. All outputs ``; inputs labelled. No our model create (watchlist loop — skipped per skill).

Driver lines:

• Cathode volume: +~3–4%/yr to ~2.4Mt, but at $0 TC/RC the processing engine contributes ~0 gross profit in all scenarios — volume growth is not an EPS lever here.
• By-product engine (the actual lever): gold ~118–139t × price; sulphuric acid 6.5Mt × price. This is where EPS is made or lost.
• Self-mined copper (~200kt): benefits from high copper price (~$10–12k/t) — a modest positive offset.
• Share count: ~3.45bn, broadly flat (no buyback/dilution culture).
• Impairment: assume a recurring ~RMB 1–1.5B drag (Aynak/offshore book).

. Cross-check: the Kristal piece's FY2028 base EPS of RMB 3.50–4.20 sits at the top of my base / into my bull — I am modestly more conservative on the base because I assume gold mean-reverts toward $3,500 rather than holding $4,000+, and that the impairment drag recurs.

Valuation translation. At HKD 30.92 (≈ RMB 28.8) and base FY26E ~RMB 2.10, the H-share trades at ~11–13.7x forward — cheap vs Western miners, fair-to-cheap vs Zijin, and the 4.4% yield is covered ~2x. The base case is not demanding; the bull case (FY28 RMB 4.2) at even a 12x multiple implies ~RMB 50 / ~HKD 54 — roughly the prior high. Brier forecast: not logged (watchlist breadth mode).

Bull vs Bear

Bull case. Jiangxi is the cheapest scaled way to own China's gold + copper + sulphuric-acid complex, mispriced because the market still files it under "troubled smelter." Its world-largest, lowest-cost smelter is a survivor of the very capacity shake-out (China cutting >10% of smelting output ) that is killing its sub-scale rivals — and when marginal smelters close, TC/RC eventually normalises in the survivors' favour, handing Jiangxi a free call on processing-margin recovery on top of the by-product earnings. Add a 4.4% covered dividend, a secular copper-deficit tailwind (S&P/Morgan Stanley/JPM all flag a 2026 deficit; AI data-centre + electrification demand to 2040 ), and a self-mined-copper kicker, and you have value + yield + optionality. The contrarian re-rating (identity catching up to earnings) is half-done, not over.

Bear case (permanent-impairment risks).

1. By-product reversion is an earnings cliff, not a wobble. With ~3.4% gross margin and ~90% of profit currently by-product-derived, gold back to $3,000 and acid to $50 takes EPS down ~30%+ and the thesis with it — the company has no processing-margin floor to catch it at $0 TC/RC. The stock is a levered gold bet that can de-rate and see earnings fall simultaneously.
2. The processing engine may never recover. If Chinese smelting overcapacity proves sticky (state-directed capacity rarely closes cleanly), TC/RC could stay ≤0 indefinitely — permanently impairing the core business's economics.
3. SOE governance / capital misallocation. Aynak (17 years, no copper), offshore stakes, RMB 1.5B impairments, no minority-protection backstop — value can leak to national objectives.

Pre-mortem (it's end-2027, the thesis broke): Gold fell back through $3,200 as real rates rose; sulphuric-acid prices normalised to ~$55; Chinese smelter capacity didn't actually cut (SOEs kept furnaces lit for employment), so TC/RC stayed negative; a fresh offshore impairment hit the book. EPS halved to ~RMB 1.3, the 4.4% yield got cut to "protect the balance sheet," and the H-share round-tripped to the high-teens — back toward its 52-week low of 14.98. The market re-filed it as "what it always was: a zero-margin toller."

Are multiples too high? No — ~11.8x trailing / ~11–13.7x forward on the H-share is not the problem. The risk is the E, not the P: the multiple is cheap precisely because the earnings are commodity-peak and arguably unsustainable. The contrarian view the market is refusing to see cuts both ways: bulls won't admit the smelting core is permanently broken; bears won't admit the by-product engine is a structural, China-policy-supported gold/acid franchise that can persist for years.

Devil's Advocate (short-seller)

Dismantling the bull case:

• The "accidental gold company" is the bulls' tell that they're paying peak-cycle earnings and calling it value. You are buying a ~1.3% net-margin toller at the exact moment two volatile commodity prices (gold, sulphuric acid) are at or near all-time highs, and pretending the resulting EPS is a run-rate. It is not. This is deep operating leverage masquerading as a moat.
• Revenue concentration is on the input side and it is lethal. ~90% purchased concentrate means Jiangxi's existence depends on access to feed at a toll that is already $0 and negative on spot. The miners hold the whip. A smelter whose core conversion margin is structurally zero is, charitably, a by-product-recovery utility — value it as such.
• The most dangerous thing bulls underestimate: Beijing. The same state that controls Jiangxi can (a) force furnaces to stay open for employment even at a loss (killing the TC/RC-recovery thesis), (b) lean on the gold price / capital controls, or (c) direct NPV-negative offshore "strategic" deals. You do not control the controlling shareholder, and its objectives are not your returns.
• Capital allocation is a graveyard of strategic bets: Aynak (17 years, zero copper, live impairment risk), recurring RMB 1.5B write-downs, illiquid offshore minority stakes. Related-party concentrate/financing with the SASAC parent is opaque by Western standards and unaudited by the SEC.
• What must hold for today's price: gold ≥ ~$3,500, sulphuric acid ≥ ~$80, copper ≥ ~$10k, and no fresh offshore impairment, and the dividend intact. That is four commodity/policy bets stacked — break any one and a ~30% EPS hit follows.
• If growth/by-product prices disappoint 20–30%: EPS to ~RMB 1.3–1.5, the yield gets cut, and on a re-rated 9–10x bear multiple the H-share has ~40%+ downside toward the high-teens — its own 52-week low is 14.98.
• The single scenario that permanently impairs: a sustained gold bear market (real-rate normalisation) coincident with permanent Chinese smelting overcapacity — both engines off at once, with no mine-quality franchise (it's only ~8–10% self-sufficient) to fall back on. Plausibility: moderate — not a base case, but very much a live tail.

Management Questions (ordered by information value)

1. With 2026 benchmark TC/RC at $0 and spot negative, what is the steady-state gross profit of the smelting/processing segment alone, stripped of by-product credits — and at what TC/RC does it turn cash-negative?
2. Quantify the by-product contribution to gross profit: at FY2025 prices, what % of group gross profit came from gold, silver and sulphuric acid respectively?
3. At what gold and sulphuric-acid prices does the consolidated business stop covering the dividend? (the floor everyone needs)
4. What is the true self-sufficiency ratio (self-mined contained copper ÷ refined cathode), and what is your target for it over five years?
5. Do you expect Chinese smelting capacity to actually be cut in 2026, and is Jiangxi cutting any of its own output, or holding furnaces lit?
6. What is the carrying value and impairment status of Mes Aynak and the other offshore stakes (First Quantum, SolGold) — and what triggers the next write-down?
7. How do you think about hedging gold/silver/acid to protect earnings, given how levered EPS now is to them — and did the trading-arm leadership change alter that policy?
8. What were the related-party transactions with Jiangxi Copper Corporation (the parent) by value in 2025 — concentrate, financing, services?
9. Capital-allocation priority for the next RMB 10B of free cash: dividends, domestic mine expansion, offshore resource M&A, or debt reduction — ranked?
10. With debt-to-asset at 57% and rising, what is your ceiling, and how do rising metal-inventory financing needs affect it?
11. What is the plan to raise mine output (Dexing and beyond) to reduce dependence on purchased concentrate?
12. How should minority H-shareholders interpret the ~50% discount to the A-share — any path to narrowing it (buyback, dual-class arbitrage, listing changes)?
13. What is your exposure to copper demand from AI data centres and the grid specifically, and how do you size that vs traditional construction demand?
14. What environmental remediation / litigation liabilities (Dexing and others) are provisioned vs contingent?
15. If gold returned to $3,000 tomorrow, what is the first lever you pull — cost, dividend, capexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs., or volume?

---