Sulphur Market Softens on Buyer Resistance Despite Persistent Supply Risk

On the same day Argus reported that China and Indonesia had gone quiet in the spot sulphur market, the US Navy confirmed it had struck Iranian rocket launchers in the Strait of Hormuz, and Iran retaliated against targets in the UAE, Jordan and then Kuwait. Vancouver fell $125 a tonne. The Strait that supposedly explains most of this year’s price action is now hosting an exchange of fire that had been paused for a month.

Three layers, and only one of them just loosened

Shanghai Metals Market described the first half of 2026 as a “three-layered squeeze”: a physical cutoff at the Strait of Hormuz that removed roughly half of the world’s seaborne sulphur trade at a stroke, a policy lockdown as Russia, Kazakhstan and Turkey each closed their own export routes within days of each other in late June, and a capacity collapse as war-damaged Middle Eastern gas and refining plants were slow to restart. That framing is useful, and it still holds, but only two of the three layers have moved since.

The policy layer has loosened. Russia’s Federal Agency for Railway Transport lifted its rail-transit ban on Kazakh sulphur on 24 July, an order that had been in place since 24 May and had forced Kazakh exporters onto costlier routes through Georgia at close to double the normal logistics cost. Kazakhstan’s own export suspension, imposed under Ministry of Energy Order No. 1363 from 27 June with an exception only for Russia-bound cargo, is understood by Argus to have been lifted in late August. By 29 August the tanker Atlantic Prestige had departed Ust-Luga carrying what Argus describes as believed to be the first Kazakh sulphur cargo since both restrictions came off, bound for Jorf Lasfar with an expected arrival around 8 September. Around 100,000 tonnes has reportedly been committed for September export.

I want to flag something in that sentence that matters more than it looks. Argus uses the word “believed.” Not confirmed. Even the trade press covering the reopening of this corridor cannot yet say with certainty what a specific cargo contains. That is not a criticism of the reporting. It is a fair description of how much visibility currently exists into a route that was closed for two months and is being pieced back together in real time.

The physical layer, meanwhile, has not loosened. It has gotten worse. The ceasefire that had held loosely since 29 July broke down at the end of August. NPR reported US strikes on Iranian rocket-launching positions inside the Strait itself, with Iranian retaliation against US-linked targets in the UAE and Jordan, the first exchange of fire in a month. By 4 September, the day after this Argus issue published, Iran had struck targets in Kuwait, extending the conflict’s geography beyond the Strait for the first time in weeks. The International Maritime Organization puts the toll at 19 seafarers killed and roughly 6,000 still stranded aboard vessels in the Persian Gulf six months into the conflict, with over 70 recorded attacks on commercial shipping. On 25 August, US officials said Navy divers and contractors had cleared more than 100 suspected mines from the Strait’s traffic separation scheme since the war began, which tells you as much about how contested the waterway still is as any price assessment does.

So the two layers are moving in opposite directions at the same time. Kazakh sulphur is coming back online through a northern route that bypasses Hormuz entirely. The Hormuz route itself is, if anything, less predictable than it was a month ago. A market that treated these as one undifferentiated “Middle East supply risk” story would be badly mispricing at least one of them right now.

Physical supply is not tradable supply

Kpler’s vessel-tracking work on this crisis makes a point that I think gets lost in most coverage: a producer having sulphur is not the same thing as the global spot market being able to buy it. By June, Kpler had confirmed that the UAE was the only Persian Gulf country still moving sulphur through the Strait at all, tracking three UAE-origin transits that month. But the same analysis carries a caveat that matters more than the headline number. Qatari, UAE and Kazakh tonnes are substantially committed under long-term contracts to India and Morocco. Even a full reopening of Hormuz would not release all of that supply into the spot market, because a meaningful share of it is already spoken for.

This is the distinction the whole market needs to hold onto right now, and it is easy to collapse without noticing. There is physical supply, meaning sulphur that exists and could theoretically be produced or recovered. There is tradable supply, meaning sulphur that a seller can actually offer into the open market rather than deliver against an existing contract. There is affordable supply, meaning tradable sulphur priced at a level some buyer is actually willing to pay right now. And there is needed supply, meaning the volume a plant requires to keep running at its intended rate. A number that looks bearish on one axis, China’s port inventory for instance, can be entirely uninformative about another.

Morocco’s OCP is the cleanest illustration of what happens when tradable supply disappears for one buyer specifically. Kpler’s tracking shows OCP received its last Persian Gulf cargo, the vessel Kallone, at Jorf Lasfar on 10 April, and has been reliant on Russian and Kazakh supply since. That matters because Kazakhstan alone supplies OCP with roughly 2.5 million tonnes a year, close to 44 percent of its total sulphur imports, according to SMM’s analysis of the trade. When Kazakhstan’s export order shut that channel on 27 June, it did not create a generic global shortage for OCP. It removed nearly half of one specific buyer’s specific supply line, and that buyer happens to be the single largest phosphate exporter on earth. The North Africa Post reported that OCP responded by pulling forward scheduled maintenance and cutting production rather than paying up for high-cost replacement cargo, essentially choosing to shrink output rather than chase spot tonnes it could not justify. That is a rational response to a supply problem. It is a completely different decision from what a buyer facing an affordability problem makes, which is to keep buying but at reduced volumes or delayed timing. Distinguishing the two matters enormously if you are trying to work out whether OCP’s next move is to buy aggressively the moment Kazakh cargo resumes, or to keep production capped regardless of price because the maintenance schedule is already locked in.

What the Chinese numbers are actually measuring

China’s port inventory fell to 963,000 tonnes as of 28 August, against 2.367 million tonnes a year earlier and 2.633 million tonnes two years earlier. Read on its own, that looks like the textbook definition of a shortage. January to July imports were down roughly 59 percent year on year in the earlier report, and that decline has continued.

But the price in the same report softened. Ex-works China eased from a recent high near Yn8,950 to Yn7,600-7,800, and cfr China granular came off $75 to $1,000-1,050. Argus attributes this explicitly to a lack of prompt buying interest rather than to any improvement in supply, noting that despite low stocks and ongoing disruptions, buyers have rejected recent price levels. China and Indonesia were reported entirely absent from the spot market in the first week of September.

This is the affordability layer taking over from the availability layer inside a single country. The mechanism that makes it possible is Sinopec and PetroChina’s discounted domestic supply programme, which Argus now reports at Yn1,750 a tonne, with up to 7 million tonnes understood to be allocated to key phosphate producers this year. That is a fifth of the open spot price. Industry sources discussed at a September meeting in Hubei whether to extend this discounted allocation through next year, alongside a separate proposal to remove domestic guidance prices for smelter acid altogether and let suppliers negotiate directly with buyers, with China’s suspension of sulphuric acid exports potentially being lifted next year subject to a floor price mechanism. None of this was formally decided as of the report date, but the direction of travel tells you something: the producers who matter most to Chinese food security are being deliberately walled off from the international benchmark, while everyone else absorbs the full swing.

There is a second demand story inside China that gets far less attention than fertilizer, and it should not be ignored. China’s expansion of lithium-iron-phosphate battery capacity has created a genuinely new sulphur demand stream, distinct from the traditional phosphate-fertilizer base, according to reporting that traced the surge in domestic prices back to a combination of collapsed imports and this newer industrial pull. Phosphate-based fertilizer is described as rigid, non-discretionary consumption tied to grain production. Battery-grade phosphate is a second, unrelated source of structural demand growth layered on top of it, at the exact moment seaborne imports fell by more than half. I would not overstate the size of this relative to fertilizer demand without better disaggregated data, but it is a real mechanism, not a speculative one, and it is one more reason China’s sulphur appetite over the next two years may not track its historical fertilizer-only baseline.

The most interesting China development this year is not a price at all. Sources told Reuters in June that the Dalian Commodity Exchange is preparing to list China’s first sulphur futures contract, potentially in the fourth quarter, explicitly accelerated by the volatility this war has caused. Domestic spot prices had spiked to a record 11,850 yuan a tonne before easing as the Strait partially reopened, a move of close to 300 percent from the prior year. Analysts at Huatai Futures framed the rationale plainly: hedging tools and visible commercial inventory at the exchange would reinforce resilience and increase China’s pricing power in the global sulphur trade. Read that phrase again. Pricing power. China imports over half its sulphur, and over half of those imports have historically come from the Middle East. A country that consumes at that scale and has just lived through a 300 percent domestic price swing has an obvious commercial incentive to stop being a price-taker on a benchmark it does not control. If this contract actually launches on schedule, it will not eliminate physical scarcity, but it will change who sets the reference price the next time a Hormuz-scale disruption happens, and that is a structural shift worth watching regardless of where spot prices sit in any given week.

The refinery connection nobody prices directly

Sulphur is a byproduct. Nobody runs a refinery harder because sulphur prices went up. Recovery is a function of crude throughput and, more specifically, of how sour that crude is, and both of those are driven by energy economics that have nothing to do with the sulphur market.

The clearest version of this right now sits in the US Gulf. EIA data show four-week average Gulf Coast refinery utilization running at 97.7 percent through 14 August, and crude imports from Venezuela averaging 630,000 barrels a day in June, up from just 86,000 barrels a year earlier, rising further to roughly 680,000 barrels a day for the four weeks to 28 August. Heavier, sourer Venezuelan crude processed at high utilization should mechanically produce more recovered sulphur. Alberta tells a similar story on the Canadian side: sulphur production in the province rose 2 percent year on year for January to July, according to the Alberta Energy Regulator, split between a 1.6 percent rise from sour gas processing and a 1.5 percent rise from oil sands facilities, both supported by strong prices for heavy sour crude and rising LNG Canada operations near Kitimat.

None of this is a coincidence of geography. It is the mechanism by which a war six thousand miles from Alberta or the Gulf coast quietly increases North American sulphur output, because higher prices for heavy sour crude make it more attractive to process, and processing more of it recovers more sulphur as a side effect. But the same mechanism cuts the other way in Russia. Gazprom’s Astrakhan gas field, with 4.8 million tonnes a year of sulphur capacity, is reportedly running only one production line, and the Orenburg facility, at 1.55 million tonnes a year, was damaged by conflict-related strikes in late June, sharply reducing Russian domestic output at the exact moment Russia’s own export ban would otherwise have made more material available to third countries. This is worth sitting with for a second: Russia, a country most people would assume to be a straightforward sulphur exporter, reportedly imported sulphur for the first time in years in October 2025, buying 35,000 tonnes at $390 a tonne, described by one analysis as marking the end of its sulphur self-sufficiency. A country that used to define part of the Black Sea supply base is now, at least occasionally, a buyer.

Japan’s refinery run rates rose 2.1 percentage points to 77.6 percent for the week to 29 August, but three crude distillation units are currently offline for maintenance or unplanned shutdown, including Eneos’s Mizushima unit closed on 31 August and its Negishi unit due offline in September, alongside Idemitsu’s Toa Oil Keihin unit shut since March. Japan and South Korea supply China with molten sulphur as a regional balancing mechanism, not a substitute for large granular cargoes, and reduced refinery activity in both countries compounds a Chinese import shortfall that already has more than one cause. None of these refinery stories move fast or make headlines on their own. Collectively they are one of the more reliable leading indicators available for where recovered sulphur volumes go next, and they deserve more attention than the freight and vessel data that dominates weekly commentary.

Phosphates: where the sulphur price stops being a sulphur problem

There is a specific point where an elevated input cost stops being absorbed and starts changing what a plant actually produces, and the financial disclosures from this year’s second quarter show that point has already been passed at more than one major producer.

Nutrien reported a negative $25 million gross margin in its phosphate segment for the second quarter, against a positive $33 million a year earlier, a swing described by the company itself as reflecting unsustainable pressure from sulphur costs on producer margins industry-wide. ICL reported the same direction, with sharply higher sulphur costs eroding phosphate margins even as higher selling prices supported revenue, and the company noted that elevated prices in Asia were already reducing what farmers were willing to apply to their fields. Mosaic has cut domestic US phosphate output by an estimated 2 million tonnes because of elevated costs, according to Kpler’s tracking. In India, GSFC’s management said on its 13 August call that average purchased sulphur costs had run $800 to $900 a tonne against a domestic spot price above 115,000 rupees a tonne, and that the company was shifting its production basket away from products with negative contribution rather than simply raising prices to match cost.

That last behaviour, changing what gets made rather than shutting the plant, is a genuinely different response from either continuing to buy at any price or stopping production entirely, and it is worth treating as its own category. A plant that idles a line is destroyed demand in the most direct sense. A plant that shifts its product mix toward whatever formulation still carries a positive margin is not gone from the sulphur market, but its consumption pattern has changed in a way that a simple operating-rate number will not capture.

At the other end of the spectrum, Argus’s own phosphate commentary for early September describes India as syphoning off global DAP supply, booking 230,000 tonnes combined via tenders and spot deals in a single week, sourced from the US, Russia, Saudi Arabia and Morocco under an offtake agreement, while Bangladesh mulls tender offers and Brazil sits largely disinterested west of Suez. That is not a market where demand has collapsed. It is a market where demand has concentrated into whichever buyers can still make the economics work, mostly subsidised south Asian markets entering their fourth-quarter application season, while everyone else is priced or timed out. Argus’s own 30-60 day view captures the asymmetry directly: DAP is likely to stay supported by that south Asian pull even as MAP, TSP and SSP prices in Brazil could slide further on fatigue and softer demand.

So has the recent easing in sulphur prices actually restored phosphate economics. Not obviously, and not yet. A negative gross margin in a single quarter does not reverse the moment the raw material cost comes off its peak, because contracts, inventory positions and prior purchasing decisions all lag the spot number by weeks or months. The more honest answer is that sulphur costs are still a floor under phosphate fertilizer prices, as Argus itself notes for the 30 to 60 day outlook, even while the sulphur benchmark itself has started to soften.

Acid, HPAL and the buyers fertilizer coverage tends to miss

Elemental sulphur and sulphuric acid are not separate markets. They are two forms of the same molecule competing for the same underlying feedstock, and treating them as unrelated misses a large and growing share of actual demand.

China’s decision to suspend exports of ordinary industrial sulphuric acid, effective from 1 May, cut exports from roughly 116,700 tonnes in May to about 980 tonnes in June, a fall of over 99 percent, with only electronic-grade acid still movable and only under special approval. That single policy decision removed a flexible substitute for elemental sulphur from the international market almost overnight, right as Indonesian nickel processors were absorbing the bulk of the physical shortage on the sulphur side. SMM’s analysis frames Indonesia’s HPAL sector, along with Morocco’s OCP and Brazil, as one of the “gravitational centres” of this crisis: Morocco as the floor-price buyer defending its phosphate empire, Brazil as the high-price buyer supporting the market’s top, and Indonesia’s HPAL demand as what the same analysis calls the rigid floor beneath high prices, describing Indonesian sulphur imports as up more than 40 percent year on year to nearly 4.8 million tonnes in the first eleven months of 2025 alone, sourced overwhelmingly from Saudi Arabia, Qatar and the UAE.

Chile’s copper industry sits on the acid side of this same competition. Reported CFR spot prices for acid into Mejillones roughly doubled between late February and mid-April, reaching $380 a tonne, and Codelco has said Middle East-related cost increases have added at least 10 cents a pound to its cash costs, with Antofagasta’s unit costs up 23 percent year on year to $3.03 a pound. Zambia introduced a permanent export licensing regime for acid at the end of March, and the Democratic Republic of Congo saw delivered prices reach $1,000 to $1,400 a tonne in the same window. None of this shows up in a weekly elemental sulphur price table, but it is drawing on exactly the same acid-production capacity that fertilizer producers are competing for, and it is one reason China’s decision to wall off its acid exports landed with more force than a purely fertilizer-focused reading would suggest.

Freight is not a footnote, it is where the arbitrage actually breaks

The freight enquiries in the latest Argus issue tell a quiet story about where alternative routes are forming. Fifteen to seventeen thousand tonnes of bulk sulphur out of Poti, Georgia, for Ashdod, Israel, with arrival windows in the 5-10 September range. Thirty thousand tonnes out of Batumi, also Georgia, for Aqaba, Jordan, arriving 10-15 September. Seven to eight thousand tonnes of bagged sulphur from Yanbu, Saudi Arabia, routed via the Red Sea to Sokhna, Egypt, rather than through the Strait at all. These are small cargoes by the standards of a market that used to move 50,000 to 60,000 tonne parcels from Vancouver to China, but they are evidence of buyers and traders deliberately routing around the two most contested waterways in the world, the Strait of Hormuz and the Black Sea, rather than paying whatever premium those routes now demand.

The Black Sea side of that calculation got materially worse over the same period this article covers. A maritime-security monitoring note dated 4 September 2026 records Novorossiysk cargo war-risk premiums at roughly 1 to 2 percent of vessel value, up from about 0.2 percent just three weeks earlier, and describes the threat as widening geographically rather than staying confined to isolated port incidents, with merchant vessels now fitted with anti-drone netting and water tanks as standard operating precaution rather than emergency measure. Ust-Luga itself, the Russian Baltic port through which the reopened Kazakh sulphur corridor now runs, has seen fires from conflict-related damage on at least two recent occasions, most recently over the weekend of 30-31 August and again on 1 September, though the European Sulphur Terminal there is understood to still be operating normally for sulphur and fertilizer exports. That is the same corridor carrying the Atlantic Prestige cargo that everyone is watching as the test case for whether Kazakh supply genuinely returns to the market. A war-risk spike at the loading port for that exact route is not a detail to file away for later.

Meanwhile the plain freight-arbitrage math from earlier this year still holds and is worth restating properly rather than as a throwaway comparison. Vancouver to China costs $33-37 a tonne for a 50,000-60,000 tonne cargo. Middle East to south China costs $140-145 for a similar parcel. At the FOB midpoints, that still put Middle East sulphur landing in China cheaper than Vancouver sulphur even before this week’s price moves, and freight was working strongly in Vancouver’s favour rather than against it. The premium buyers pay for Vancouver material is not explained by shipping cost. It is explained by Sultran’s dual-terminal loading structure at Pacific Coast Terminals and Vancouver Wharves, which lets the port keep exporting even when one facility is down for scheduled maintenance, as Pacific Coast Terminals was between 9 and 22 September, or when wildfire disruption threatens the rail corridor through Boston Bar, as it did earlier this year before conditions began stabilising. A buyer paying the Vancouver premium is buying a lower probability of the cargo arriving late, not a lower freight bill.

Producers are not selling into a single market-clearing price

A softer spot assessment does not mean every producer accepted a lower price for the same reason. Argus’s own reporting this week shows a Canadian cargo concluded at a $950 a tonne FOB netback, $125 below the prior assessed range, described explicitly as product struggling to find a buyer, alongside separate, unrelated business: a copper-belt-linked sale to Namibia rejected at $1,200 a tonne, an Indian sale concluded at $1,050, and Brazilian buyers relying on existing stock rather than fresh purchases at all.

Those are not four data points converging on one number. They are four different sellers making four different decisions about volume versus margin, in four buyer relationships with different urgency levels. The Canadian netback sale looks like a producer choosing to move volume rather than hold inventory against a market that is not currently paying up. The rejected Namibian offer looks like a buyer that has already secured what it needs for the period and is not willing to chase price regardless of supplier posture. Reading either of these as “the new market price” and extrapolating forward is the same mistake as reading China’s port inventory number without asking whether that inventory is actually available to the buyer trying to source it.

Demand destruction and demand postponement are not the same thing, and the difference matters more now than it did in July

For most of this year, the useful distinction was between buyers who could afford current prices and buyers who could not. That framing is starting to break down, because a second distinction has opened up underneath it: buyers who have permanently reduced consumption, and buyers who have simply timed their purchases differently while waiting for clarity.

Nutrien’s negative phosphate margin and Mosaic’s 2-million-tonne output cut look like the first category, actual reduced production tied to unworkable economics. GSFC’s shift in product mix sits somewhere in between, reduced consumption of specific negative-margin products rather than a blanket production cut. India’s DAP buying spree in early September, alongside its still-open granular sulphur tender that received a single offer in mid-August and has not yet been awarded, looks like the second category: a buyer whose underlying need has not gone anywhere and who is actively purchasing where the economics work, while waiting on a proposed subsidy mechanism before committing fully elsewhere.

This distinction matters because it changes what happens if China’s port inventory starts drawing down again or if the Kazakh corridor proves less reliable than currently hoped. Destroyed demand does not snap back quickly. Postponed demand can return all at once, and a market that has priced in the first kind of buyer disappearing permanently can be badly wrong-footed if it turns out to have been the second kind waiting for a specific trigger, a subsidy approval, a tender award, a cargo actually discharging on schedule.

What I would actually watch over the next 30 to 60 days

Whether the Atlantic Prestige and the cargoes following it into Jorf Lasfar actually discharge on the schedule currently expected, and whether the volume genuinely reflects Kazakh-origin material rather than a mixed or uncertain cargo, given that even Argus is only willing to say it is “believed” to be the first such shipment.

Whether the ceasefire that broke down at the end of August stabilises again or continues expanding geographically, given that Iran has now struck targets in Kuwait as well as the UAE and Jordan within the space of a week.

Whether China’s Dalian Commodity Exchange actually lists a sulphur futures contract in the fourth quarter as currently planned, and what that does to the visibility of Chinese demand and inventory once a paper market exists alongside the physical one.

Whether the discounted Sinopec and PetroChina supply programme is formally extended into next year, and whether China’s suspended sulphuric acid export regime reopens under a floor-price mechanism, both of which were under discussion but undecided as of early September.

Whether Novorossiysk and Ust-Luga war-risk premiums keep climbing at the rate seen over the past three weeks, and what that does to the cost of moving Kazakh and Russian sulphur through the Baltic and Black Sea corridors that the market is currently counting on as the primary non-Hormuz alternative.

Whether India’s cabinet actually approves the proposed sulphur and sulphuric acid subsidy, and whether tender participation, currently thin enough that a bulk sulphur tender drew a single offer, changes materially once that policy question is resolved one way or the other.

None of these are questions a price chart can answer on its own. They require watching vessels, ports, refinery utilization, corporate disclosures and a war that is not currently behaving the way a six-month-old conflict usually does, which is fading. It is not fading. It escalated again in the same week the price of the commodity it disrupted went down. Anyone trading, buying or producing sulphur right now should probably sit with that fact for longer than the price table invites them to.