From Shortage to Surplus: Why This Alumina Cycle Reversed So Quickly

High profits last year spurred the entire industry to rush into capacity expansion. This year, new domestic capacity, imported supplies, and Guinean ore have all converged simultaneously—why can’t alumina maintain profitability like copper does?

—— ◆ ——

Core Summary · Alumina prices surged to around 5,500 RMB/ton in 2024, but by the first half of 2026 had fallen below total cost, with the industry broadly facing per-ton losses · In 2025, actual new domestic capacity reached 8.3 million tons; an additional 10–13 million tons are planned for 2026, with 60% concentrated in Q2, primarily in Guangxi · By late June, net imports replaced net exports; port and plant ore inventories combined totaled approximately 93 million tons, sufficient for 4–5 months · Mineral Cube Assessment: The supply response speed of alumina is far faster than that of copper mines, which is the fundamental reason it cannot replicate the long-term high-profit model seen in copper

Trigger I: Current Alumina Status – Prices Clinging to Cost Lines, Inventories Continuing to Accumulate

As of the end of July 2026, China’s installed alumina capacity has reached 120.95 million tons/year, with operational capacity at 95.9 million tons/year; August production is expected to exceed 8 million tons. Meanwhile, total domestic alumina inventory stood at 6.398 million tons, an increase of 138,000 tons from the previous month, continuing its upward trend. The reference trading range for the main futures contract is 2,600 to 2,800 RMB/ton, while tender and procurement prices in Xinjiang and other regions were approximately 2,985 to 2,990 RMB/ton (delivered price). Industry theoretical profits fluctuated narrowly between a loss and a profit of 300 RMB/ton, leaving little room for comfort. Viewed over the past six months, this is not a brief price correction, but a systemic oversupply lasting more than half a year, permeating both the supply side and the cost structure.

This scene is almost worlds apart from what existed just over a year ago. In 2024, driven by disruptions in Guinean ore supplies, alumina prices once spiked to around 5,500 RMB/ton, opening up significant profit margins for the industry. Almost all major players made the same decision simultaneously: expand capacity. This is the starting point for understanding today’s oversupply situation.

The economies of scale brought about by high profits continue to self-reinforce: In Q1 2026, the capacity utilization rate of major domestic alumina enterprises generally remained above 88%, with leading companies such as Weiqiao Pioneering Group and Jinjiang Group approaching 100%. Unit fixed asset depreciation and management expenses were further diluted, continuously expanding the cost advantages of leading enterprises. This also means that smaller capacities face maintenance and production cut pressures earlier during price downturns.

Background II: How Last Year’s High Profits Turned Into This Year’s Widespread Losses

In 2025, China’s total alumina capacity reached 108 million tons/year, with actual new capacity coming online totaling 8.3 million tons, mainly concentrated in the first half of the year. Total domestic alumina production reached 94.448 million tons, a year-on-year increase of 9.4%, of which metallurgical-grade output was 89.28 million tons. Overseas metallurgical-grade alumina production stood at 60.62 million tons, an incremental increase of approximately 2.54 million tons. The concentrated release of supply quickly crushed prices: although the industry’s theoretical loss narrowed to within 100 RMB/ton by the end of 2025, the sector looked entirely different from the high-profit state at the beginning of the year.

In 2026, the pace of capacity release has not slowed: Domestic planned new capacity ranges from 11.6 to 13 million tons, with 60% concentrated in Q2, making Guangxi the absolute main battlefield. On April 21, a newly built alumina plant in Guangxi ignited its calcination furnace and began producing products on the same day, involving a capacity of 1.2 million tons/year. Internationally, Indonesia is expected to export 5 million tons of alumina to China in 2026, continuously impacting import windows with its cost advantages. In April, China’s alumina imports surged by 86.9% month-on-month. The excess domestic supply was so great that even export channels struggled to absorb it—from a net export of 860,000 tons in the same period last year to a net import pattern from January to April, forcing excess pressure to be digested internally within the domestic market.

Longer-term capacity layouts are also advancing: Chinalco Hong Kong, a subsidiary of Chinalco, plans to invest in and build a 1.2-million-ton alumina project in Guinea, with a total investment of approximately $1 billion. In July 2026, relevant agreements for the project were approved by the President of Guinea. On May 21, Chinalco Hong Kong formally signed the project agreement with the Guinean government in Conakry. This marks another major layout by Chinese capital in Guinea following the Simandou iron ore project. Once these projects come online, they will extend alumina supply increments further into Guinea itself, further reducing the possibility of the industry regaining pricing power in the medium to long term. Such capital layouts are not isolated cases; multiple Chinese enterprises have been simultaneously advancing integrated mine and alumina projects in Guinea in recent years, meaning the inertia of supply-side expansion will likely continue for the next few years.

Breakdown III: From Ore to Alumina – How Oversupply Transmitted Downstream Step by Step

The source of alumina oversupply must be traced one level higher—the bauxite ore itself is also experiencing oversupply. Global new bauxite supply in 2025 amounted to approximately 25 million tons (Guinea 16 million tons, China 3 million tons, Indonesia Nanshan 3 million tons, non-mainstream sources 3 million tons), whereas the ore increment corresponding to new alumina demand was only about 11.6 million tons. The ore supply increment was more than double the demand increment. In 2026, Guinea plans to add another 25 million tons of bauxite, and the CTC project in Sierra Leone has the potential for an additional 10 million tons, making ore-side oversupply nearly a certainty.

TimeCapacity/ProductionPriceInventory/Imports & Exports
2024 PeakProfit-driven industry-wide expansionSpot prices surged to approx. 5,500 RMB/tonLow inventory
Full Year 2025Total capacity 108 million tons, +8.3 million tons new (concentrated H1)Margins narrowed; year-end theoretical loss <100 RMB/tonConcentrated supply release; inventory turned upward
April 2026Installed domestic capacity approx. 115 million tonsSpot price 2,680 RMB/ton, falling below total cost of approx. 2,760 RMB/tonSocial inventory 5.887 million tons, historical peak; Imports +86.9% MoM in April
June 2026Planned new capacity 11.6–13 million tons, 60% concentrated in Q2Futures touched lows near 3,300 RMB/tonNet import pattern replaced year-ago net exports from Jan–Apr
End of July 2026Installed capacity 120.95 million tons, operational capacity 95.9 million tonsMain contract range 2,600–2,800 RMB/tonTotal inventory 6.398 million tons, +138,000 tons MoM
August 2026Operational capacity rises to approx. 97.7 million tons; production to break 8 million tonsAustralia FOB approx. $362.5/ton (transaction price Aug 14)Xinjiang delivered price 2,985–2,990 RMB/ton, declining MoM

Changes in futures warehouse receipts further amplified the visibility of this oversupply: In late January 2026, alumina warehouse receipts were approximately 120,000 tons; by late March, they had increased to over 400,000 tons, an increment of nearly 300,000 tons. This was primarily due to arbitrage opportunities between futures and spot markets created by the oversupply, combined with the off-season consumption, leading to a large volume of inventory being pushed onto the exchange floor. Based on an expiration cycle of approximately five months, a large number of receipts are expected to expire centrally starting in July and August, which is also one of the reasons for the sustained pressure on prices in the third quarter.

Demand Side Offers Few Surprises

The demand side also presents few surprises: Electrolytic aluminum accounts for over 90% of downstream demand for alumina. However, by 2026, the capacity utilization rate of electrolytic aluminum is already as high as 97%, approaching its "ceiling," which corresponds to an alumina demand growth rate of less than 2%. Non-metallurgical demand (chemicals, electronics, new energy materials), while growing at a rate of 8% to 10%, represents only about 8% of the total volume. This scale is insufficient to offset the low elasticity of mainstream demand. Consequently, it is unrealistic to expect the excess supply to be naturally absorbed by the demand side; instead, contraction must rely primarily on the supply side itself.

Lesson 4: The Key Difference Between Alumina and Copper: A One-Order-of-Magnitude Gap in Supply Response Speed

Both are resource commodities, and both have experienced periods of high profitability, yet alumina and copper have followed entirely different scripts. The root cause lies in supply elasticity. Statistics show that for the world’s largest 35 copper mines, the average time from discovery to production startup is 16.9 years. In contrast, new alumina capacity typically requires only one to two years from planning to ignition and production startup. For example, the 1.2 million-ton alumina plant in Guangxi that ignited in April had a window of merely a few months between publicly disclosed construction progress and the official release of products.

The disparity in resource constraints is equally stark: Global exploration investment in copper declined from $16.59 billion during 2011–2015 to $9.44 billion during 2016–2020, a drop of 43.1%. Although investment rebounded slightly during 2021–2023, only one large-scale new copper deposit was discovered globally during that same period. Conversely, Guinea alone holds bauxite reserves of 7.4 billion tons, accounting for 24% of the global total and ranking first worldwide; raw material constraints are virtually non-existent. Furthermore, the Bayer process used for alumina production is highly standardized. As long as capital, electricity, and caustic soda supplies are available, capacity can be rapidly replicated. This stands in complete contrast to copper mining, which relies on geological discoveries and requires years of community negotiations and environmental approvals. This explains why the wave of high profits near 5,500 RMB/ton in 2024 was filled by industry-wide expansion impulses in just over a year, whereas even when copper prices reached historical highs, global mine output growth remained constrained to around 1% to 2%.

Comparison 5: Alumina vs. Copper: Identical High-Profit Signals, Two Different Supply Responses

DimensionAluminaCopper Mines
Typical Construction Cycle1–2 years (planning to production)Average 16.9 years for the top 35 global mines (discovery to production)
Raw Material / Resource ConstraintsGuinea alone holds 7.4 billion tons of reserves; supply is dispersed and continuously expandingGlobal reserve growth has slowed; reserves fell 2% YoY in 2024; only one large deposit discovered in the past three years
Trend in Exploration/Capital Investment (Past Three Years)Capacity expansion relies mainly on replicating existing processes, with relatively low entry thresholdsExploration investment dropped 43.1% compared to the previous five-year period (2016–2020), constraining future supply
Supply Response Following High ProfitsConcentrated expansion within 1–2 years, causing prices to quickly fall back near cost linesEven with record-high prices, global mine output growth has remained in the 1%–2% range for many years

Differences in industry organizational structure further amplify this gap in response speed: The Chinese alumina industry features numerous participants. Companies such as Weiqiao, Jinjiang, Chinalco, East Hope, and Xinfeng all possess the financial and engineering capabilities to rapidly launch new production lines. Industry concentration is not particularly high, meaning any single company can independently decide to expand capacity upon seeing a profit window. In contrast, global copper resources are highly concentrated in a few countries like Chile and Peru, and among a handful of multinational mining companies such as BHP, Freeport-McMoan, and Anglo American. The capital expenditure for a single copper project often runs into billions of dollars, and decision-making cycles are measured in years, making it impossible to ignite and start production within months as seen with alumina plants. These differences in equity and resource distribution structures fundamentally determine the vast disparity in the speed at which the supply side can respond to identical high-profit signals in these two industries.

This comparison also explains why the projected growth rate of global copper mine output was revised down from an initial 2.3% to 1.4% for 2025–2026, leaving a final global nominal surplus of only 178,000 tons. Meanwhile, the alumina surplus scale during the same period reached 5 million tons, with a surplus rate nearing 5%. Both sectors faced robust demand, but the answers provided by the supply side were completely different.

Assessment 6: Kuang Lifang’s Judgment: This Round of Excess Is Structural, Not Short-Term

Kuang Lifang’s assessment is that unless there are unexpected disruptions at the bauxite source in Guinea, or if large-scale mandatory de-capacity measures are implemented domestically, the alumina industry is unlikely to return to a state of widespread profitability within 2026. The reason is straightforward: as long as prices offer a profit margin of even 1,000 to 2,000 RMB, new capacity will follow to fill the gap within one to two years. This response speed is rare among resource-based industries. Looking at a longer time horizon, this is not the first time alumina has experienced such a cycle: every window of high profitability is exchanged for the release of capacity on a larger base in the next round, and the industry seems unable to break out of this loop.

Variables worth monitoring are instead on the cost side: Guinea’s export quota policies, trends in ocean freight rates, and the maintenance schedules of high-cost domestic capacity. Any unexpected tightening in any of these links could pull industry profits back from losses toward breakeven faster than a "demand revival." However, given the current capacity base, it is difficult to replicate the high profits of 5,500 RMB/ton seen in 2024.

It is important to additionally highlight the price differential signal between overseas and domestic markets: On August 14, alumina traded on the overseas market at an Australian FOB price of $362.5 per ton. Converted at the prevailing exchange rate, this is nearly identical to the domestic delivered price. This indicates that the excess supply is not a situation unique to China; the restart of overseas smelting capacity is also continuously absorbing alumina supply globally. Relying on export channels to significantly digest domestic excess capacity is unrealistic in the short term.

Outlook 7: Three Recommendations for Readers

Rather than guessing when alumina prices will rebound, readers should focus their attention on specific variables on the supply side, as the lessons from this round of the cycle are clear enough: expectations on the demand side have never been the primary variable determining prices in this industry; rather, it is the pace of supply-side expansion. Here are three specific recommendations:

  • Recommendation 1: Set the actual production startup pace of domestic new capacity (rather than planning announcements) as the core tracking indicator. Pay particular attention to ignition progress in the Guangxi region, as this is the primary signal for judging whether the scale of excess supply continues to expand.
  • Recommendation 2: Simultaneously track Guinea’s export quota policies and trends in ocean freight rates. Tightening in either of these variables could alter the logic of cost support faster than domestic production cuts.
  • Recommendation 3: Do not apply the supply logic of copper mines to alumina. History repeatedly shows that in the alumina industry, high profits are often digested by expansion impulses before the following year arrives. If similar price spikes occur in the next round, akin to 2024, it is likely to follow the same script.

—— ◆ ——

Alumina profits are never kept. It is not that the industry lacks effort, but rather that its supply curve is inherently faster than price movements.

— END —