Tianqi Lithium

Phase A — Understand the business

Lens 1 · Company Overview

Tianqi Lithium is a vertically-integrated lithium major: it mines spodumene, converts it to lithium chemicals (carbonate, hydroxide, chloride, metal), and holds two of the best lithium resource equity stakes on earth. Founded 1992; the modern company dates to founder Jiang Weiping's 2004 acquisition of Shehong Lithium; listed in Shenzhen in 2010 (002466.SZ) and Hong Kong in July 2022 (9696.HK). HQ: Chengdu, Sichuan.

The business has three profit engines, only one of which it fully controls:

1. Upstream resource (the crown jewel). A 26.01% effective interest in the Greenbushes mine in Western Australia — the world's largest and lowest-cost hard-rock (spodumene) lithium operation. Ownership chain: Talison Lithium = 51% TLEA + 49% Albemarle; TLEA (Tianqi Lithium Energy Australia) = 51% Tianqi + 49% IGO → Tianqi effective 26.01%, IGO 24.99%, Albemarle 49%. Talison's Phase III chemical-grade expansion commenced commissioning 18 Dec 2025, lifting spodumene capacity to 2.14 Mtpa — Tianqi's look-through share ≈ ~556 ktpa SC6.
2. Midstream conversion (China + Australia). ~91,600 t/yr of built lithium-chemical capacity across five sites: Shehong (24 kt: 14.5 kt carbonate / 5 kt hydroxide / 4.5 kt chloride), Zhangjiagang (20 kt battery-grade carbonate), Anju (23 kt battery-grade carbonate), Tongliang (600 t lithium metal) and Kwinana, WA (24 kt hydroxide, Train 1 — chronically under-ramped).
3. Financial resource stakes. A ~22.16% stake in SQM (Sociedad Química y Minera de Chile), the world's second-largest lithium producer (Atacama brine) — accounted for by the equity method, so SQM's earnings swing straight into Tianqi's P&L. Plus 20% of the Zhabuye salt lake (Shigatse, Tibet).

Customers: Chinese and global cathode/battery makers (CATL, BYD-chain, LG, etc. — Tianqi sells largely into the merchant lithium-chemical market with a mix of long-term floor-price contracts and spot). Suppliers: effectively itself upstream (Greenbushes offtake) plus the Cuola/lepidolite domestic feed. Competitors: Albemarle, SQM, Ganfeng Lithium, Pilbara Minerals, Mineral Resources, plus a long tail of Chinese lepidolite/salt-lake converters. Contract structure: battery makers and Tianqi have negotiated floor-price mechanisms averaging ~$18,000/t in long-term supply agreements, but the marginal tonne prices off a violently cyclical spot market.

Lens 2 · Supply Chain — named stakeholders

Upstream → midstream → end market, names attached:

• Ore (spodumene): Greenbushes (Talison; co-owners Tianqi/IGO/Albemarle) is the anchor feed — hard-rock, ~2.14 Mtpa SC6, lowest cash cost in the industry. Secondary domestic feed: Yajiang Cuola spodumene (Sichuan; wholly-owned, ~19.71 Mt ore ≈ 630 kt LCE, undeveloped — climate/altitude/transport) and lepidolite.
• Brine (equity, not offtake): SQM (Salar de Atacama, Chile — 22.16%) and Zhabuye (Tibet — 20%). These deliver earnings, not tonnes Tianqi controls.
• Conversion: Tianqi plants at Shehong, Zhangjiagang, Anju, Tongliang (China) and Kwinana (Australia).
• Midstream partners / co-owners: IGO Limited (49% of the Australian TLEA JV — funder of the 2020 bailout), Albemarle (49% of Talison, and Greenbushes' other offtaker).
• End customers: cathode-active-material makers and cell manufacturers — CATL, BYD, LG Energy Solution, Samsung SDI, Panasonic-chain, feeding EV OEMs and (increasingly) grid-scale energy storage.

Chokepoints / single-source dependencies:

• Greenbushes is the single point of concentration — it is the asset. A disruption there (or an adverse offtake-pricing reset) hits the whole model. The 2024 collapse was largely a Greenbushes-offtake pricing-mismatch event (see Lens 5), not a demand event.
• Chinese conversion is not differentiated — it competes with a flooded domestic converter market; the moat is entirely upstream (Greenbushes) and in the SQM equity, not in the chemical plants.
• Kwinana is a negative node — it consumes Greenbushes concentrate at a cash loss.

Lens 3 · Competitive Advantages (moats)

The moat is a rock, not a company. Tianqi's durable advantage is ownership of the best lithium orebody in the world (Greenbushes) plus a low-cost brine call option (SQM) — assets that cannot be replicated and sit at the very bottom of the global cost curve. That is a genuine, resource-based moat: in a commodity, the low-cost producer survives every down-cycle that bankrupts the high-cost tail. Greenbushes' cash cost is widely described as the lowest in hard-rock lithium; SQM's Atacama brine sits in the "Tier 1, $3–5/kg" lowest-cost cohort.

But the moat is heavily qualified:

• Tianqi does not operate its crown jewel — it owns 26% of Greenbushes and 22% of SQM. It is a minority financial holder of both, with Albemarle/IGO controlling Greenbushes economics and the Chilean state (Codelco) now controlling SQM (Lens 5/13). Minority ownership of great assets is a weaker moat than operating them.
• No switching-cost / brand / network moat in the chemical business — lithium carbonate is a fungible commodity; Tianqi is a price-taker downstream.
• Bargaining power is asymmetric: strong over its own offtake pricing at Greenbushes (as the 2025 "pricing-mechanism optimization" showed — they fixed the mismatch in their favor), weak versus Chinese battery buyers in a surplus year.

Verdict on moat: real but rented. You are underwriting orebody quality and cost-curve position, diluted by minority stakes and a state counterparty.

Lens 4 · Segments

No segments.csv on the shelf (web-only). Directionally, from FY2025 commentary:

• Two reported product lines: lithium concentrate (upstream, Greenbushes-derived) and lithium compounds & derivatives (midstream chemicals). In FY2025 the upstream concentrate segment carried the profitability on strong margins while the downstream compound segment margins declined — Morningstar's read of FY2025: "Strong Upstream Mining Growth and Margin Offset Declining Compound Segment."
• Geography: production straddles Australia (Greenbushes concentrate, Kwinana hydroxide) and China (Shehong/Zhangjiagang/Anju/Tongliang); sales are China-weighted (battery supply chain) with global chemical offtake.
• The SQM equity stake is a de facto third "segment" running through investment income / share-of-associate — and it is the single most volatile line (−RMB ~1.7B drag in 2024 on the Chilean tax ruling; a positive contributor again in 2025 and a major driver of the Q1-2026 surge).

FY2025 total revenue RMB 10.346B, −20.8% YoY — revenue fell even as the company swung to profit, because the recovery was margin-and-associate driven, not volume-driven.

---

Phase B — Measure performance

Lens 5 · Earnings Result — the turnaround print and the boom-bust arc

The most important thing to understand about Tianqi is the shape of its earnings: it is a hyper-cyclical levered to a commodity that moved ~10× and back. Net profit attributable to shareholders:

Latest audited print (FY2025): revenue RMB 10.346B (−20.8%), net profit attrib. RMB 463M (from −RMB 7.905B), EPS RMB 0.28, weighted ROE 1.1% (+18.02pp), total assets RMB 72.11B (+5%), debt-to-asset ratio 28.04% (−0.35pp).

Three drivers of the 2025 turnaround (well-sourced, consistent across outlets):

1. Concentrate pricing-mechanism optimization — Talison's Greenbushes offtake to Tianqi was re-priced to track the market on a much shorter lag, ending the 2024 mismatch where Tianqi bought feed at lagged-high prices and sold chemicals at crashed-spot prices. This restored core margin.
2. SQM investment income recovery — the 2024 Chilean-tax hit annualized out; SQM earnings rebounded.
3. FX gains on a stronger Australian dollar (AUD-denominated Greenbushes/TLEA assets), plus lower asset impairments than 2024.

⚠ Flagged conflict (analytically important). During 2025, management guided a full-year net profit of RMB 3.69–5.53B, yet the audited actual came in at just RMB 463M. With Q1 2025 at ~RMB 104M, this implies a weak, even loss-making, back half of 2025 (heavy H2 impairments and/or a still-depressed lithium price before the Q4-2025→Q1-2026 spike). The takeaway: the recovery was thin and back-loaded until the price genuinely inflected in early 2026 — do not extrapolate the RMB 2.0B Q1-2026 run-rate naively.

Market reaction: HK shares gapped up >8% on the Q1-2026 profit alert (21 Apr 2026). Balance-sheet flags: net debt only ~RMB 4.56B (total debt 14.53B − cash 9.97B), D/E 0.26, equity RMB 56.88B — a fortress compared to the 2020 near-default. The 2024 loss did not threaten solvency; the 2020 debt did. That is the single biggest structural improvement in the story.

Lens 6 · Earnings Calls / management messaging (sentiment trend)

No transcripts on the shelf; synthesized from results briefings and profit alerts:

• 2024 (trough): contrition + capex retrenchment — "most loss-making year since listing," capacity plans revised down, Kwinana Phase II terminated as "economically unviable," Kwinana expansion suspended. Founder stepped down as chairman. Tone: defensive, survival-mode.
• 2025 (stabilizing): pivot to "pricing-mechanism alignment" and "resilient financial structure"; strategy slogan "consolidate upstream, strengthen midstream, penetrate downstream." Operating cash flow turned positive (Q1'25 ~RMB 620M).
• 2026 (re-acceleration): overtly bullish — profit alerts crediting "lithium price recovery" + "SQM investment gains"; JV CEO says Kwinana "to be competitive soon."

What they stopped saying: the 2022–23 mega-expansion / capacity-leadership rhetoric is gone. What they now repeat: resource security (advancing the domestic Cuola project), balance-sheet discipline, and pricing-mechanism control. Sentiment arc: panic (2024) → stabilization (2025) → confidence (2026), tracking the lithium price almost one-for-one — appropriate for a pure cyclical, but a reminder that management tone here is a coincident, not leading, indicator.

Lens 7 · Comps

Lithium pure-plays, market caps:

Look-through cross-check: Tianqi's 22% of SQM's $21B mkt cap ≈ ~$4.6B — i.e. ~a third of Tianqi's own ~$11–14B market value is a stake it has just announced it will sell down. That is the central valuation tension.

Lens 8 · Stock-Price Catalysts (>5% moves, ~5 years)

The tape is driven by (1) the lithium price and (2) SQM headlines — almost nothing else moves it materially:

• 2020 (down, near-death): looming $1.9B CITIC loan maturity, ~RMB 1.29B cash — solvency scare; rescued by IGO's $1.4B TLEA investment.
• 2021–Nov 2022 (up, mania): carbonate to RMB ~512k/t; 2022 profit +1,060%; HK IPO (Jul 2022).
• Nov 2022 → 2024 (down, crash): lithium −~90% from peak; stock −43%+, ~$22B HK market value erased; profit warnings, Kwinana halt, founder resignation, SQM tax hit.
• Apr 2024 (down): Chilean court tax ruling → ~$260M hit to Tianqi via SQM; biggest quarterly loss.
• 2025 (up, stabilization): return to profit after five loss quarters.
• Jan–Feb 2026 (mixed): lost final Chile Supreme Court appeal on the SQM-Codelco deal (down/overhang) → announced SQM stake trim (Feb 2026).
• Apr 2026 (up, +8% gap): Q1-2026 profit alert +15–18×.

Pattern: this is a high-beta lithium-price proxy with a Chilean-политics overlay. It does not react to product cycles, management execution, or capital returns — it reacts to spodumene/carbonate spot and to SQM's fate. Trade it as such.

---

Phase C — Judge people & books

Lens 9 · Management

• Founder Jiang Weiping (蒋卫平) — acquired Shehong Lithium (2004) and built Tianqi into a global major via two audacious M&A coups: Greenbushes/Talison (2012–13, outbidding Rockwood/Albemarle) and SQM (2018, $4.07B for 23.77% from Nutrien). 20+ years in lithium. The Greenbushes deal is one of the great resource acquisitions of the era; the SQM deal nearly bankrupted the company (Lens 5/10).
• Succession: After the record loss, Jiang Weiping stepped down as chairman in April 2024, handing to his daughter Jiang Anqi (蒋安琪, b. 1987) — vice-chair for the prior two years, ~10 years in the industry, VP/director of Tianqi Group since 2016/18. Jiang Weiping remains executive director. Former president Vivian Wu departed August 2020 at the depth of the debt crisis.
• Skin in the game / control: founder-family-controlled via Chengdu Tianqi Industry (Group). This is the single most important governance fact — a family holding company controls the listed entity; minority public holders ride alongside.
• Capital-allocation history — the crux, and it is polarized: brilliant on acquiring (Greenbushes = generational), reckless on financing it (the 2018 SQM deal was ~90% debt-funded into a cyclical top and detonated in 2020). Post-crisis discipline has been genuine: deleveraged from near-default to 28% debt/asset, terminated a value-losing expansion (Kwinana II), and is now exiting the SQM misadventure. ROE is a cyclical rollercoaster (1.1% in 2025 vs the 2022 peak), so a through-cycle ROIC read is meaningless without normalization.
• Archetype: founder-operator / empire-builder, now in a professionalizing, deleveraging phase under the second generation. Implication: expect resource-security ambition (Cuola) tempered by a hard-learned balance-sheet caution. The family will run it for control and staying power, not for minority-shareholder returns per se.

Lens 10 · Forensic Red Flags

No filings on the shelf; the accounting risk map is reasoned from public disclosure and the model's structure, and should be re-checked against the audited HKEX/CNINFO 2025 report:

• Equity-method income from SQM is the #1 accounting-quality flag — a ~22% associate whose earnings, tax disputes, and now state-restructuring flow straight into Tianqi's net income as a non-cash, non-controlled line. It produced a ~RMB 1.7B swing down in 2024 and a large swing up in 2025–26. Investment income of this magnitude, from an entity Tianqi does not control and is now selling, inflates earnings volatility and quality risk. Strip it out to see the core miner.
• Inventory / pricing-mismatch losses (2024) — the loss was amplified by carrying spodumene feed at lagged-high cost into a crashed chemical market; watch inventory NRV write-downs and the concentrate-vs-compound margin spread each half.
• Asset impairments — 2024 carried elevated impairments (Kwinana, and IGO fully impaired its 49% JV stake, ~A$605M, on a A$955M FY25 plant loss ). Kwinana remains an impairment-watch asset until it demonstrably clears cash cost.
• FX — AUD/USD/RMB translation on Australian assets drives real net-income swings (a 2025 tailwind, a potential future headwind).
• Cash-flow vs earnings — with equity-method income and FX in the mix, operating cash flow is the cleaner truth than net income; OCF turned positive in 2025 after being negative — a genuine positive, and the metric to track.

Regulatory findings (required sub-section). Per regulatory/regulatory-findings.md (generated 2026-07-10, sources SEC EDGAR EFTS LR + AAER): total_sec_findings: 0 — Tianqi has no CIK and is not an SEC filer, so no EDGAR enforcement search is possible. Non-SEC / non-US findings via web search:

• Chile (material, ongoing): Tianqi has been in protracted litigation with Chilean regulators and SQM's board over the SQM–Codelco tie-up — contesting the regulator's move, demanding a shareholder vote, and losing its final appeal at the Chilean Supreme Court (Jan 2026), which held that a mandatory shareholder vote was not required despite Tianqi's 22% holding. Tianqi keeps the door open to international arbitration. This is a shareholder-rights / minority-protection dispute, not an enforcement action against Tianqi — but it is the defining legal overhang.
• No material SEC/FTC/DOJ/FDA-type enforcement action against Tianqi was surfaced in this pass (it is not US-regulated). As a Shenzhen/HK issuer it is subject to CSRC/SEHK rules; no material sanction surfaced. Verified via SEC EDGAR EFTS (no CIK → n/a) + web search as of 2026-07-10.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection (EPS, next three fiscal years)

Bottom-up is impossible without the shelf; this is a top-down cyclical scenario set, anchored to the sourced forward multiple, and labeled `` throughout. Share count 1.641B (1,477.07M A + 164.12M H).

Market-implied anchor: forward P/E 11.76× on mkt cap RMB 80.67B ⇒ market-implied FY2026 net profit ≈ RMB 6.86B, EPS ≈ RMB 4.18. This is the consensus the tape is discounting.

FY2027–FY2028: genuinely unforecastable at single-point precision for a pure cyclical — the 2022→2024 swing (from +RMB 24B to −RMB 8B) is the honest confidence interval. Directionally: if the 2026 supply additions (per WoodMac/S&P surplus calls) hit, 2027 normalizes down from a 2026 peak; if the deficit calls (Morgan Stanley/Fastmarkets) win, EPS extends higher. Base-case through-cycle "mid" earnings power ≈ RMB 4–6B/yr (EPS ~2.4–3.7), with peaks and troughs an order of magnitude wider. No forecast logged — this is an unattended --watchlist run (skip forecast.ts create; a Brier forecast should be logged only on a genuinely committed base case in a human-gated /thesis pass).

Lens 12 · Bull vs Bear

Bull case. You are buying the world's best lithium orebody at ~1.4× book and ~12× forward earnings, into a violent up-cycle (carbonate ~doubled to ~$26k/t in Q1'26; Q1 profit +15–18×). Greenbushes' cost position means Tianqi prints cash at prices that shutter the marginal tonne; the balance sheet is now a fortress (net debt ~RMB 4.6B, D/E 0.26) so it can hold through cycles rather than dilute at the bottom; Phase-III adds low-cost volume; the SQM exit crystallizes ~$4–5B and removes a Chilean-politics albatross; and structural EV + energy-storage demand growth underwrites multi-year lithium demand. Operating + financial leverage make it the highest-beta way to be long a lithium recovery.

Bear case (permanent-impairment risks). (1) It is a price-taker on a commodity that fell ~90% — the terminal value is a bet on a spot price no one can forecast, and 2026 surplus calls (WoodMac ~109kt, S&P) are live. (2) The SQM stake — a third of look-through value — has been structurally expropriated of upside (Codelco 51% control; ~70% of Atacama earnings to the Chilean state from 2025) and Tianqi is now a forced seller into that, likely below its 2018 $4.07B cost. (3) Kwinana has never earned its cost of capital and is a standing impairment risk. Pre-mortem (18 months out, thesis broke): lithium supply (Chinese lepidolite + African + Argentine brine) outran storage-led demand, carbonate slid back toward $12–14k, the 2026 SQM windfall reversed as the Codelco structure diluted the associate line, and the stock round-tripped the 2026 rally — because you were long a cyclical at a cyclical high on a temporary SQM tailwind. Multiples: ~12× forward looks fine only if 2026 is mid-cycle; if 2026 is peak, the through-cycle multiple is far higher and the stock is expensive.

Contrarian view (what the market is refusing to see): consensus is fixated on the price-recovery beta and treating the SQM stake as an asset being monetized. The market is under-weighting that selling SQM removes Tianqi's only brine diversification and leaves it a near-pure, minority-owned bet on a single Australian rock — concentrating, not de-risking, the business — right as it hands the crown-jewel brine franchise to a state competitor. The re-rating narrative and the de-diversification are the same event.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the bull case:

• Revenue/earnings concentration: two assets Tianqi does not operate (26% Greenbushes, 22% SQM) drive the economics. Shift either — an offtake-pricing reset at Greenbushes, or the Codelco dilution at SQM — and the earnings base moves without management touching a lever.
• The moat is weaker than bulls think because minority ≠ control. Albemarle/IGO set Greenbushes' economics; Chile now controls SQM. Tianqi captures orebody quality only at others' governance mercy.
• Most dangerous competitor bulls underestimate: not Albemarle — the Chinese lepidolite + integrated converter complex (Ganfeng, CATL's own upstream, Zangge, Yongxing) that can flood domestic carbonate and cap the price Tianqi's chemical segment realizes. And African spodumene (Zimbabwe/Mali) is a new low-ish-cost tonne source.
• Worst capital-allocation move: the 2018 all-debt SQM buy at the top — a near-fatal error whose cleanup (forced stake sale at/below cost, years of interest drag, a 2020 rights issue at a low, IGO selling half the Australian unit) is still ongoing eight years later. That is the base rate for how this management deploys a windfall.
• Assumptions that must hold for today's price: lithium averages ≥ ~$18–20k/t through 2026; the SQM equity line stays positive despite Codelco; Kwinana stops bleeding. If lithium disappoints 20–30%, the base EPS roughly halves (RMB ~6.5B → ~3B) and a 12× forward multiple on a falling number de-rates twice.
• Single scenario that permanently impairs: a structural lithium surplus (Chinese lepidolite economics + African supply + storage demand undershoot) pins carbonate at $10–12k for years — Kwinana closes for good, the SQM sale prints a loss, and Tianqi becomes a ~26% stub of one mine trading at book. Plausibility: moderate — it is exactly the 2023–24 experience, and the supply side has not structurally tightened.

Lens 14 · Fifteen Questions for Management (ordered by information value)

1. Post-Codelco, what is the run-rate equity-method contribution you expect from SQM, and at what price/volume do you sell down — full exit or a held residual?
2. What is the Greenbushes offtake pricing formula now (lag, benchmark, CIF/FOB), and how does it behave in a falling price environment vs the 2024 mismatch?
3. What is Kwinana's cash cost per tonne today, and at what lithium-hydroxide price does it clear cash breakeven — and if it can't in this up-cycle, why keep it?
4. What is the use of proceeds from the SQM sale — debt paydown, buyback, Cuola development, or new M&A? (The market needs to know the empire-building reflex is retired.)
5. What is the capex, timeline, and all-in cost to bring Yajiang Cuola into production given altitude/transport, and what carbonate price justifies it?
6. Through the cycle, what normalized ROIC do you underwrite for the consolidated business ex-SQM?
7. How do you think about balance-sheet leverage at the next cyclical top — is there a hard debt ceiling after 2018–2020?
8. What share of 2026 volume is under floor-price contracts vs spot, and where are the floors?
9. What is your read on 2026–27 lithium supply-demand (surplus vs deficit), and how is the business positioned if surplus wins?
10. What is the capital-return policy (dividend/buyback) now that the balance sheet is repaired?
11. How exposed are you to AUD/USD/RMB, and do you hedge translation on the Australian assets?
12. What is the relationship and any residual governance/arbitration path with SQM/Codelco after the Supreme Court loss?
13. What is the founder family's long-term intent for control, and how are minority-holder interests protected in related-party dealings with Tianqi Group?
14. What ends the structural discount of the H-shares (9696) to the A-shares, and would you buy back H?
15. Where do you want Tianqi on the "consolidate/strengthen/penetrate" strategy in five years — a pure low-cost upstream miner, or an integrated materials player?

---