El Niño Is Official and Cocoa Is Down 34%: Soft Markets Are Pricing Inventory, Not Forecasts
NOAA’s Climate Prediction Center moved from El Niño Watch to El Niño Advisory on 11 June. The distinction is not cosmetic. A watch says an event is likely to form; an advisory says it has arrived. The June discussion put a 63% probability on a very strong event during November through January, which would put it among the largest in the record back to 1950, and a 97% chance conditions persist into early spring 2027. Every soft commodity desk on both sides of the Atlantic has been looking at that same forecast for eight weeks.
Cocoa is down roughly 34% from a year ago.
That gap is the whole story, and it is being misread. The consensus framing is that El Niño threatens the world’s most-traded agricultural commodities, which is true as a statement about crop physiology and false as a statement about what futures are currently doing. Soft markets are not paying for the forecast. They are paying for inventory they can verify, and the two numbers are pointing in opposite directions.
Take cocoa on its own terms. Ghana’s COCOBOD has guided 2026/27 production down to a range of 450,000 to 550,000 tonnes against 750,000 projected for the current season, citing swollen shoot disease, ageing plantations, illegal gold mining and the likelihood of adverse weather. Ivory Coast is expected to fall more than 10%. That is a supply guide of a severity that in most crops would be a standing bid. New York cocoa did rally hard on it, up 7.4% in a session to $5,490, and London put in a 6.9% day. Then it faded. Prices sit near $5,660 after touching an eight-month high of $6,455 on 9 July, inside a 52-week range that runs from $2,846 all the way to $8,823. Meanwhile Ivory Coast port arrivals since the October crop year opening came in 18% above the same period last year. The forecast is bearish supply. The tape is bullish supply. Traders are trading the tape. Citigroup went neutral on cocoa after having been bullish, saying explicitly that it wants to see evidence of El Niño crop damage before it goes back.
Now look at coffee, where the market behaves completely differently. Arabica is bid around $3.23 to $3.27 per pound and grinding higher. The reason is not a rainfall model. ICE certified stocks stood at 260,720 bags on 3 August, down from 292,810 a week earlier and against 760,529 bags at the same point last year, the lowest since January 2024. That is a countable number in a warehouse. Coffee gets paid because the scarcity has already happened and someone can audit it. Cocoa gets a forecast and a shrug.
Sugar sits between the two and shows the mechanism cleanly. Raw sugar pushed to a four-week high at 15.31 cents, white to $478.90, and a Reuters poll of eleven traders and analysts has the market ending the year about 4% higher as the balance flips into deficit. The near-term drivers are physical and observable: rain interrupting the Brazilian cane harvest, mills diverting a larger share of cane into ethanol than last year, heat in the EU. The El Niño component, India receiving below-average August monsoon rainfall after an average July, is still a weather service statement rather than a tonnage figure. Sugar is rallying on the first set and holding an option on the second.
History supports the discount rather than the panic. The 2015/16 El Niño was rated very strong and moved cocoa very little. The 250% cocoa move of 2023/24 is now routinely filed as the El Niño trade, but swollen shoot, aged trees, deferred replanting and Ghanaian farmgate pricing were all inside that number, and the weather was the accelerant rather than the cause. Anyone sizing a position off the 2023/24 analogue is buying a composite and calling it a single factor.
The lag structure is what most of the commentary gets wrong. West Africa dries early in an El Niño, and the test for the 2026/27 main crop is rainfall across August through October, during establishment. That window is open right now, which makes cocoa the near-term expression. Southeast Asia dries late, with palm oil yields in Malaysia and Indonesia responding six to twelve months after the event through fresh fruit bunch formation. Robusta and rubber run on a similar delay. If this event does reach very strong intensity through the northern winter, palm is a 2027 problem being discussed in 2026, and the futures curve is not being asked to price it yet.
One framing point worth holding onto. The World Bank has headline commodity prices rising about 16% in 2026, the first annual increase since 2022, and that increase is energy and fertiliser. Agricultural prices fall in the baseline. El Niño is therefore the upside risk to a bearish forecast, not the forecast itself. Those are very different positions to carry, and they size differently.
The advisory is eight weeks old. The rain gauges in Ivory Coast will settle it by November.