The Market Is Pricing the Fire, Not the Tank Farm

Key takeaway: Malaysian palm oil futures reached their highest level since December 2024 on Thursday before settling at 4,898 ringgit a tonne on Friday, up 1.74% on the week. The near-term balance is loose, with end-September stocks expected near 3 million tonnes and 1-15 September exports down 17.8% to 25.6% on the month. The rally is about 2027, not October: Indonesia is fighting its worst wildfires in 11 years, and Kalimantan output is projected to fall 12% to 15% in the fourth quarter after an estimated 11% to 12% drop in August.

The benchmark December contract on Bursa Malaysia printed a 21-month high on Thursday morning, traded almost unchanged at 4,996 ringgit by the midday break, and then gave back 38 ringgit on Friday to close at 4,898 ringgit, or about $1,200 a tonne. That is a 0.77% daily loss inside a 1.74% weekly gain. Rival oils dragged on the last session: the most active Dalian soyoil contract fell 1.57%, Dalian palm lost 2.14% and Chicago soyoil slipped 0.77%. Crude oil fell 2% for a third straight day as worries over Saudi supply eased, and a firmer ringgit added a second headwind for dollar-based buyers.

None of that is the story. The story is that palm made a new cycle high in a week when every fundamental a trader can see on a spreadsheet said it should not have.

A loose front end

Start with Malaysia. Cargo surveyors put exports of Malaysian palm products for 1-15 September between 17.8% and 25.6% below the same period of August. Production is in its seasonal peak. Traders expect end-September inventories to reach around 3 million tonnes, a level that would ordinarily cap any rally. Malaysia's palm oil council has already flagged a strong El Niño as a 2027 output risk rather than a 2026 one, which puts the physical shortage a year away while the stock build is here now.

Friday had an extra ingredient. Traders in Kuala Lumpur are waiting for an official word from New Delhi on a possible cut to India's edible oil import duty, and spent the week trading the rumour rather than the fact.

A very tight back end

Now look at Indonesia, which is where the price is coming from.

The country is dealing with its most severe fire season in 11 years, with forest and peatland dried out by El Niño conditions. Government figures show roughly 202,000 hectares burned between January and July. An environmental group estimates a further 600,000 hectares were damaged in August alone as the fires intensified. Central and West Kalimantan, the provinces hit hardest, produce about 13% and 14% of Indonesia's palm oil respectively. Add East Kalimantan and the three provinces account for around 36% of national output.

An agricultural commodities analyst told Reuters on Wednesday that Kalimantan output likely fell 11% to 12% in August and could fall 12% to 15% in October to December after two very dry months. GAPKI, the Indonesian producers' association, has cut its 2026 national output estimate by 2.9% because of the fires. The association's chairman noted that fire-damaged palms can die or suffer stress and take about three years to recover.

Run the arithmetic on the analyst's numbers. If a third of Indonesia's supply loses 12% to 15% in the fourth quarter, the national output hit is in the region of 4% to 5% for the period. That is our calculation, not a forecast from any official body, and it comes on top of a 2027 El Niño threat to Malaysia and a B50 biodiesel mandate that has been absorbing Indonesian palm since July. GAPKI still expects 2026 output to exceed 2025 as new estates mature. The point is that the growth cushion the market had been counting on has been cut, and the damage has a three-year tail.

Why the curve is doing what it is doing

Put the two halves together and the price behaviour makes sense. The front of the palm curve is trading the tank farm: heavy Malaysian stocks, thin exports, softer crude. The back of the curve is trading the fire: lower Indonesian availability through year-end, structural biodiesel absorption, and a producing region that cannot replant its way out of the damage before 2029. A market that hits a 21-month high on the day its exports are down a quarter is a market whose marginal buyer is not looking at this month. Add the talk that Jakarta could lift the blend rate beyond B50, which would divert still more palm from export tanks into domestic fuel, and the back end has every reason to stay firm.

What this means for India

India bought 782,761 tonnes of palm oil in August, the most since February, and refiners went into September carrying heavy port stocks. That looked like good timing when it was done. It looks better now. The forward supply Indian refiners will need for the post-Diwali quarter is the same forward supply Kalimantan will not deliver.

Three things follow. The relative value trade that pulled Indian demand toward soyoil in August is likely to persist, because soyoil's supply story is a record South American crop while palm's is a fire. Any Indian duty cut lands on a palm market that has lost its production growth for the year, so the pass-through into origin prices will be quicker than in 2025. And the seasonal Malaysian stock build that usually hands Indian buyers a fourth-quarter discount may be the last cheap window before the El Niño year. The market has already started to charge for it.

The number to watch is not next month's MPOB stock figure, which will look bearish. It is the December output data out of Indonesia, and whether the 12% to 15% Kalimantan decline shows up in the national total. If it does, 4,996 ringgit will not be the high.

Supply shocks that take three years to heal change the conversation about long-term sourcing, replanting economics and the true cost of biodiesel mandates. That conversation is the centrepiece of GLOBOIL India 2026, the 29th edition, running 29 September to 1 October 2026 at The Westin Mumbai Powai Lake, Mumbai. The world's leading edible oil and agri-trade conference brings Indonesian and Malaysian producers, Indian refiners and the analysts tracking the fire damage into the same room, one week before the fourth-quarter numbers start to print.

Sources (internal, remove before publishing)

• https://www.brecorder.com/news/40440142/palm-oil-falls-on-weaker-rival-oils-crude-gains-17-for-the-week (Reuters, 18 Sep 2026: Dec contract close 4,898 ringgit, -0.77%, +1.74% w/w; Dalian/CBOT moves; crude -2%; ringgit +0.49%; India duty watch)
• https://www.brecorder.com/news/40439934/palm-retreats-from-21-month-high-on-stock-outlook-weak-exports (Reuters, 17 Sep 2026: highest since Dec 2024; midday 4,996 ringgit; end-Sept stocks ~3 Mt; 1-15 Sep exports -17.8% to -25.6%)
• https://www.brecorder.com/news/40439822/palm-oil-output-in-major-indonesian-production-region-likely-to-fall (Reuters, 17 Sep 2026: Kalimantan Q4 -12% to -15%, Aug -11% to -12%; worst fires in 11 years; 202,000 ha Jan-Jul, ~600,000 ha Aug; province shares; GAPKI -2.9%; 3-year recovery)
• https://www.brecorder.com/news/40439591/indias-august-soyoil-palm-oil-imports-rise-on-stocking (Reuters/SEA, 15 Sep 2026: India Aug palm imports 782,761 t, highest since Feb)
• https://www.palmoilmagazine.com/hot-news/2026/08/31/super-el-nino-threatens-malaysia-palm-oil-output-in-2027-mpoc-sees-monsoon-as-cushion/ (MPOC on 2027 El Niño risk)