The squeeze's anatomy
A 64% inventory collapse, a $323 backwardation, and a Chinese export wave that arrived just in time
The squeeze built in three acts. Between December 2024 and mid-August 2026, LME warehouse stocks collapsed 64% from roughly 264,000 tonnes to 88,000 tonnes. The cash premium over three-month metal exploded to $323/ton in October 2025 — the highest level since at least 1997 — before easing to $215/ton in late August 2026. Three-month LME zinc hit a four-year high of $4,107/ton cash-settled on Aug. 27, 2026, up 55% from a mid-2025 trough near $2,650.
LME zinc: from squeeze to relief
LME stocks vs. backwardation: how the squeeze peaked, then broke
Unit: tonnes
Dec 2024 stocks
264,000
Mid-Aug 2026 stocks
88,000
Oct 6, 2026 stocks
126,975
Apr 2026 stocks
98,650
Backwardation blew out, then narrowed
Cash premium over three-month LME zinc, $/tonne
Unit: $/t
Oct 2025
323
Late Aug 2026
215
Sept 2, 2026
139
Oct 6, 2026
60
Then China opened the tap. August exports of 40,668 tonnes were the second-largest monthly total in nearly two decades, lifting LME inventory to a four-month high and pulling the three-month price back from $4,065/ton in early September to $3,766/ton on Oct. 6, 2026. Hong Kong took 35,000 tonnes since mid-August, Singapore 21,250 tonnes and Kaohsiung 16,725 tonnes — three LME delivery points absorbing 80% of August outbound shipments.
The mechanism
Why China suddenly had metal to spare — and why the West needed it
China became a net refined-zinc exporter of 4,388 tonnes over the first eight months of 2026, on course for its first annual net-export year since 2022. The driver was a textbook arbitrage: the LME backwardation pulled Chinese metal into nearby Asian warehouses faster than customs data — released on a seven-week lag — could capture it. Through August, China ran with refined zinc output up 5.9% YoY while the rest of the world contracted 3.4% YoY in the first half.
- Chinese refined output rose 5.9% YoY in H1 2026 while Western output fell 3.4% YoY — a ~9-point divergence that left China holding the only sizable spare metal (ILZSG).
- Zinc-concentrate imports into China ran +30% YoY in 2025 and a further +5% in the first seven months of 2026, feeding Chinese smelters that the West's mines can no longer supply to their own plants.
- Spot treatment charges for zinc concentrate imported into China collapsed to -$113/ton in August 2026 — an all-time low, against the $85/ton benchmark Teck and Korea Zinc settled earlier in the year.
- Negative TCs mean smelters now pay miners to take ore rather than earn a processing fee — the strongest signal that the Western smelting base is structurally short, not just temporarily squeezed.
The structural gap
Western mines are not coming back, and smelters can't run without concentrate
The ILZSG swung its 2026 zinc balance by 300,000 tonnes — from a forecast 271,000-tonne surplus to a 29,000-tonne deficit — almost entirely because of Western mine and smelter attrition. Global mine output shrank 2.6% YoY in the first half, and the cumulative decline from 2015 to 2025 has already taken 8.6% out of global production.
| Producer | Asset / event | 2026 zinc impact | Status |
|---|---|---|---|
| Teck Resources | Red Dog (Alaska) | Zinc-in-concentrate guidance cut to 375,000–415,000 tonnes (from 430,000–480,000) | Mine life to 2031; closure risk by 2032 |
| Teck Resources | Antamina (Peru, JV) | Zinc guidance cut to 35,000–45,000 tonnes (from 55,000–65,000) | Lower-zinc ore; structural decline |
| Glencore | Lady Loretta (Australia) | H1 2026 own-source zinc fell 21% YoY to 365,600 tonnes | Mine reached end of life in 2025 |
| Glencore | Kazzinc (Kazakhstan) | Ust-Kamenogorsk zinc-dust incident, May 2026 | Guidance unchanged; further risk to H2 |
| Nexa Resources | Cerro Lindo (Peru) | Q2 2026 zinc 79 kt, +8% YoY on grades | Block-cave start July 2026; 7% production cut guided for 2026 |
| Nyrstar (Trafigura subsidiary) | Budel (Netherlands) | Strategic review launched Sept 24, 2026 | Capacity 315,000 t/yr; at-risk |
Trafigura's footprint made the squeeze worse. In early August 2026, a single short position equivalent to 20–29% of open interest in the LME August zinc contract sat on the books as the cash premium held $60/ton above three-month metal. With Nyrstar's Budel smelter — 315,000-tonne annual capacity — now under formal review and Nyrstar CEO Guido Janssen publicly warning the Dutch plant is \"losing money every year,\" the Western smelting base is the part of the chain that cannot bounce when prices rise.
The demand angle no one is talking about
AI racks, power grids, and the 50%-of-demand hiding inside galvanized steel
More than half of global zinc consumption runs through galvanizing — and the AI build-out is the fastest-growing slice. Hot-dip galvanized steel is the standard for power-generation and transmission infrastructure, switchgear enclosures, cable management and data-center structural steel. Nucor now claims 95% of all steel products that go into a data center, with management citing \"a few examples\" of HSS structural tubing, rebar and galvanized product lines built specifically for that customer base. Construction accounts for more than half of global zinc usage, and the data-center construction market is on track for 7.5–8.1% CAGR through 2030.
- Zinc die-casting is the dominant process for data-center power, cooling and networking enclosures, server-rack chassis, heat sinks, and connectors — zinc's EMI shielding and heat-dissipation properties make it the default for high-density AI infrastructure (Dynacast, CWM Diecast).
- The AI server chassis market alone is forecast to reach USD 12.6 billion by 2035 at 18.5% CAGR from a 2026 base, per industry sizing, with zinc die-cast chassis competing against aluminum for cooling-intensive 1 MW rack designs.
- China's die-cast zinc-alloy operating rate fell to 38.18% in August 2026 — down 5.22 points YoY — as auto and property weakness capped domestic demand and freed Chinese metal for export to LME warehouses.
- Galvanized-steel demand for renewable energy, grid hardening, and data-center power is on a multi-year structural ramp independent of the property cycle, leaving zinc consumption tied to the AI capex line, not just Chinese construction.
What the tape already says
Zinc-exposed equities are up 40–60% YTD — but smelters are still trapped
Equity markets have already priced the squeeze in the miners. Year-to-date through early October 2026, Teck Resources is up roughly 47%, Glencore is up 46%, and Nexa Resources is up 60%. Teck Resources delivered Q2 2026 revenue of CAD 3.61 billion with net income of CAD 854 million, and Q1 2026 revenue of CAD 3.94 billion with net income of CAD 819 million — both quarters well above year-ago levels. Nexa Resources reported Q2 2026 net income of US$98 million on US$899.7 million of revenue, with mining zinc production of 79 kt, up 8% YoY on higher grades.
LME zinc stocks (Oct 6, 2026)
126,975 t
Highest since June 2025; up 44% from mid-Aug low
LME 3-month zinc (Oct 6, 2026)
$3,766/t
Down from $4,065/t early September
Cash premium over 3-month (Oct 6)
$60/t
From $215/t in late August
China refined zinc exports, Aug 2026
40,668 t
2nd-highest monthly in ~20 years
2026 ILZSG balance swing
300,000 t into deficit
From 271k t surplus to 29k t deficit
Spot zinc concentrate TC (China)
-$113/t
All-time low; benchmark $85/t
| Company | Mkt cap | TTM revenue | YTD share move | Verdict |
|---|---|---|---|---|
| Teck Resources | CAD 46.05B | CAD 14.0B (TTM) | ≈ +47% | Pure zinc exposure via Red Dog; mine life ends 2031 |
| Glencore | US$89.7B | US$304.7B (TTM) | ≈ +46% | Diversified; H1 2026 own-source zinc -21% YoY |
| Nexa Resources | US$1.65B | US$3.44B (TTM) | ≈ +60% | Mid-cap pure-play; Cerro Lindo silver tailwind Q2 2026 |
| POSCO Holdings | KRW 23.5T (ADR) | — | Mixed | Galvanized steel buyer; passes zinc cost into steel |
| Nucor | US$57.2B | — | Strong YTD | Galvanized steel seller; 95% of data-center steel |
| Steel Dynamics | US$34.1B | — | Strong YTD | EAF flat-rolled; galvanized capacity exposure |
What breaks the squeeze
Three triggers that re-open the arbitrage — and one that doesn't
The squeeze's relief valve sits in three places: China, demand, and Western smelters. First, China. If domestic property or auto demand recovers — China's die-cast zinc-alloy operating rate sits at 38.18% versus prior years above 50% — China will absorb its own metal and stop exporting. Second, demand destruction. China auto sales are already down 22% YoY; global growth softness would accelerate that and kill the squeeze from below. Third, smelter restart economics. Spot TCs need to flip back to positive (above the $85/ton benchmark) before Nyrstar's Budel and similar capacity come back online, and that requires either a concentrate supply response or a smelter rationalization that hasn't started.
- Short-term (days–quarters): A drop in Chinese refined-zinc exports below ~25,000 tonnes/month, combined with LME stocks falling below 88,000 tonnes, would re-open the path to $4,065/ton and beyond — the cash premium's floor sits near $60/ton.
- Medium-term (1–2 years): The Anglo Teck merger (awaiting only Chinese regulatory approval, expected by March 2027) creates Anglo Teck — combining Teck's Red Dog and Antamina with Anglo's Collahuasi and Quellaveco copper — and reshapes the Western critical-minerals supply base alongside zinc.
- Long-term (1–3 years): Red Dog's scheduled 2031 closure removes ~5% of global zinc supply with no large-scale replacement in development; Glencore's Lady Loretta is already gone, Nexa Resources's Cerro Lindo grades are declining, and Chinese smelters are the only capacity expanding.
- Risk to the thesis: A faster-than-expected concentrate supply response — new mines in Latin America or Africa — would flip TCs positive and reopen the Western smelting base, breaking the structural part of the shortage. Until then, miners capture the rent, smelters pay to operate, and China arbitrages the gap.
Investable names
- Pure-play zinc via Red Dog (375,000–415,000 t 2026 guidance) and Antamina; Q2 2026 net income of CAD 854M confirms high-price flow-through.
- Red Dog mine life ends 2031; this scarcity window is one of the last cycles for tier-one Western zinc before structural supply tightens further.
- Merger of equals with Anglo American awaits only Chinese approval (expected by March 2027) — adds copper but does not dilute the zinc-cycle exposure through the deal-close horizon.
- Own-source zinc fell 21% YoY to 365,600 t in H1 2026 as Lady Loretta ended; remaining McArthur River output still monetizes the high-price window.
- Trading arm booked EBITDA of ~US$3.3B in H1 2026; the firm likely held the LME August short that drove the squeeze, lifting marketing EBIT further.
- Diversified copper-coal-recycling mix dilutes the zinc-leverage thesis versus Teck or Nexa, but provides downside cushion if zinc rolls over.
- Smaller-cap pure-play with +60% YTD share move already; Q2 2026 net income of US$98M (zinc 79 kt +8% YoY) validates margin capture on the squeeze.
- Cerro Lindo silver-stream obligation falls from 65% to 25% from Q2 2026, freeing roughly 40% of by-product silver for spot-price sales through 2027.
- Block-cave start at Cerro Lindo (July 2026) is the operational thesis: 7% production cut in 2026 transitions to a new cost curve from 2027 onward.
- Supplies 95% of the steel in a data center — HSS tubing, rebar, galvanized structural — making it the cleanest AI-grid trade with implicit zinc-price exposure.
- Galvanized steel margins benefit from tight supply and AI/data-center demand, but higher zinc costs flow into COGS, capping the upside.
- Market cap of US$57.2B (up from US$37.6B at FY25 close) reflects a substantial re-rating already; the next leg needs hot-rolled coil pricing to follow.
- EAF flat-rolled producer with growing galvanized and painted-steel capacity; benefits when AI/data-center demand pulls coated-steel volumes.
- Market cap of US$34.1B (up from US$25.0B at FY25 close) shows the re-rating is partly priced; remaining upside ties to coated-product mix.
- Net debt/EBITDA at ~1.2× leaves balance-sheet headroom for further coated-line capex if data-center order book extends into 2027.
- Korean steelmaker sells into the same galvanized-steel demand pool as Nucor; merger of Korea Zinc and Young Poong in 2026 changes its zinc procurement cost structure.
- Korea Zinc's domestic M&A battle (Young Poong/MBK controlling bid) is a binary catalyst to watch through year-end 2026 for downstream galvanizing cost dynamics.
- ADR trades at ~KRW 23.5T market cap, well below 2023 levels; re-rating depends on Chinese steel export discipline and Korean mill utilization.
- Listed Chinese zinc smelter and miner; directly benefits from the Chinese export wave now earning arbitrage on LME-back prices.
- Negative TCs in China (-$113/ton) still favor integrated smelters with captive concentrate, and Yunnan Chihong holds both mining and smelting capacity.
- Germanium and lead by-product exposure diversifies earnings away from pure zinc price moves through 2027.
- Shanghai-listed zinc smelter operating at scale inside China's 5.9%-YoY output growth band; benefits while domestic operating rate remains below 40% and exports dominate.
- Arbitrage economics (LME backwardation over SHFE) underwrite realized prices 5–10% above domestic benchmarks for export-bound tonnes.
- P/E near 10× on recent prints leaves room for re-rating if China's net-export year extends into 2027, even at moderating LME prices.
