Antimony Price Discovery Cracks as Four Benchmarks Slip

Fastmarkets stumbled this week in its assessment of one of the world’s most geopolitically sensitive critical minerals. A procedural error delayed four global antimony trioxide benchmarks, arriving alongside a set of methodology clarifications that expose just how fragile pricing can be in thinly traded markets.

The delays landed on 9 September 2026, when four trioxide assessments spanning China, Rotterdam, Antwerp, and Baltimore published late. They came just days after Fastmarkets clarified how it normalises data for its Rotterdam warehouse grades, a technical detail that quietly reshapes which trades qualify as benchmark-setting.

For anyone whose contracts index off these codes, the timing matters. Antimony pricing already moves violently under Chinese export controls, and a benchmark hiccup in a market this opaque is not a rounding error.

This piece breaks down which assessments face delays or corrections, how the grade normalisation rules are shifting, and how you can protect contract settlements from the reporting inaccuracies that thin markets invite.

The 9 September trioxide benchmark disruptions

Four global antimony trioxide benchmarks published late on 9 September 2026 after a procedural error inside Fastmarkets’ assessment process. The agency subsequently updated its pricing database to reflect the affected numbers, but the delay left a window where automated data feeds carried stale or missing figures.

The affected assessments stretch across the physical hubs that anchor global antimony trade:

• MB-SB-0006: Antimony trioxide 99.5% Sb₂O₃ minimum, FOB China, priced in USD per tonne
• MB-SB-0004: Antimony trioxide 99.5% Sb₂O₃ minimum, CIF Antwerp/Rotterdam, priced in USD per tonne
• MB-SB-0009: Antimony trioxide 99.5% Sb₂O₃ minimum, in-warehouse Antwerp/Rotterdam, priced in EUR per kg
• MB-SB-0010: Antimony trioxide 99.5% Sb₂O₃ minimum, in-warehouse Baltimore, priced in USD per lb

That geographic spread matters. A single procedural slip touched the Chinese export price, the European import and warehouse quotes, and the US warehouse reference simultaneously, meaning almost no major trioxide trade route escaped the disruption.

Fastmarkets keeps a channel open for stakeholders to push back. Traders and buyers can submit feedback or corrected price data through the agency’s designated email addresses, referencing antimony trioxide in the subject line. Comments flagged as confidential are treated that way, while others may be shared on request.

Here is the practical read for you. If your supply contracts settle against any of these four codes, do not trust an automated feed this week. Pull the figures directly from the updated pricing database and verify them by hand before signing off on any pending cargo settlement.

For investors indexing off-take agreements to these benchmarks, the friction is short-lived but real. Knowing exactly which hubs were hit lets you audit exposed cargoes fast rather than discovering a settlement mismatch after the fact.

Grade normalisation and the Rotterdam warehouse clarifications

The delays were only half the story. On 8 September 2026, Fastmarkets issued a separate clarification on how it handles data submission and normalisation for two Rotterdam warehouse metal assessments, and the detail reveals just how much subjective judgment sits behind a published antimony price.

Normalisation is the process by which reporters adjust trade data that does not exactly match a benchmark specification, correcting for timing, lot size, or impurity levels, so it can be compared against on-spec material. In the clarification, Fastmarkets reiterated its hierarchy: trades that outright meet the specification rank higher than any adjusted data, with warehouse material qualifying only if lot-size timing falls within 10 calendar days.

The antimony trioxide price assessment methodology sits at the centre of this problem, because the normalisation rules that determine which trades qualify as benchmark-setting are less visible to contract holders than the final published figure.

The two grades in question look similar on paper but serve different buyers. MB-SB-0001 carries a tight bismuth limit, while MB-SB-0002 follows the broader Minor Metals Trade Association (MMTA) standard grade II, which prioritises arsenic and selenium limits over bismuth.

Around the same date, Fastmarkets also corrected the MB-SB-0001 assessment originally published on 5 August 2026. The revised range was widened to fully incorporate a reported deal at USD 18,000 per tonne that had not been captured in the first publication.

What this tells you is uncomfortable but useful: benchmark prices are not raw market snapshots. They are curated figures, shaped by which deals qualify and how off-spec trades get adjusted upward or downward toward the reference grade.

That is why the price you see published can feel disconnected from the physical quotes your suppliers actually send. Understanding the normalisation hierarchy explains the gap, and it tells you to treat a single benchmark print as an assessment, not gospel.

Geopolitical pricing pressure and illiquid market risks

Strip away the administrative detail and a bigger problem emerges. These are not just clerical stumbles; they are symptoms of a market where price discovery rests on a handful of verifiable trades, all subject to state control.

Antimony has been one of the most violent critical minerals of the decade. When China imposed its US-specific export ban in December 2024, Rotterdam prices surged toward USD 39,500 to 40,000 per tonne, roughly a 250% increase over the year. From a 2019 baseline near USD 5,500 per tonne, the metal peaked at USD 57,000 to 63,000 per tonne in mid-2025.

Chinese export controls have been the dominant structural force behind antimony’s price trajectory since late 2024, with each policy update reshaping the pool of licensed trades that reporting agencies can use to anchor their assessments.

In a market that thin, a single distorted data point carries outsized weight. Physical spot liquidity is low, so one delayed trade or one questionable submission can swing the assessed price, and that number then becomes the reference point for every subsequent negotiation.

China’s whitelist system and the pool of verifiable trades

The structural risk sharpens when you consider who controls supply. Chinese mining and refining dominate global antimony, and over 2026 Beijing built a state-trading whitelist permitting only 11 companies to export the metal, mirroring similar regimes for tungsten and silver.

A whitelist system means price discovery depends on a very small pool of licensed, verifiable trades. When the physical market is deliberately managed for volume and timing, the reporting agency has fewer genuine transactions to anchor its assessment, amplifying the impact of any single error.

S&P Global has warned that boutique materials like antimony lack exchange-cleared benchmarks, leaving agency assessments as effectively the only point of truth. You should read these delays accordingly: not as isolated mistakes, but as evidence of a structurally opaque market where one bad data point can distort your portfolio valuation. For anyone modelling revenues on new non-Chinese antimony projects, pricing opacity is a variable you cannot afford to ignore.

Protecting critical mineral offtake agreements from single point failures

The lesson from this week is not that Fastmarkets erred. It is that leaning on any single index in an opaque market leaves you exposed to a correction you never saw coming.

The divergence is documented. For critical minerals, assessments from agencies such as Fastmarkets, Benchmark Mineral Intelligence, and Asian Metal can differ by 10 to 30%, driven by differing specifications, geography, and access to transaction data. Relying on one series in a contract formula invites material mispricing.

Regulators have flagged the gap directly.

The FCA review of benchmark-sector controls, published in July 2025, found that gaps in control frameworks raise the risk of inaccurate benchmarks being produced, with potential harm to end-users and reputational damage to administrators.

The UK Financial Conduct Authority’s July 2025 review of benchmark-sector controls found that weaknesses in control frameworks raise the risk of inaccurate benchmarks, with potential harm to end-users and reputational damage to administrators.

Fastmarkets itself acknowledges the limit. Its methodology states that evaluations may not conform to prices available from third parties, and that users rely on published figures at their own risk. When a benchmark evaluation fails to match physical reality, the consequences can escalate into margin calls or litigation over settlement values.

The practical step is to build cross-checking into your contracts. Mandate multi-source price discovery, combining agency benchmarks with customs data, procurement indices, and a second reporting agency, rather than trusting a single potentially flawed index.

Offtake agreement risk management in other critical mineral markets, particularly copper concentrate, has generated a set of contract structures and price-averaging clauses that antimony buyers can adapt to insulate settlements from single-index failures.

As new Western supply slowly comes online, pricing transparency should improve, but only gradually. Until then, insulate your agreements now.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results, and financial projections are subject to market conditions and various risk factors.