El Niño’s Direction Is Settled. Its Size Is Not

El Niño’s emergence is now near-certain; its peak strength is the variable that prices into 2027, and no agency will yet call it.

Joao Morciani · Senior Agricultural Analyst

Our CEO, Francisco Martin-Rayo, described three clocks running at different speeds. Two have now produced their answers; the third — the climate clock — is the one that resolves over the next ninety days. This week, we got the first piece of the answer.

We have tracked these clocks in sequence, most recently through the biofuels-as-strategic-asset reframe, repricing the oil complex. The climate clock is the one we turn to now, and the answer it just produced is partial in a revealing way. The latest National Oceanic and Atmospheric Administration (NOAA) Climate Prediction Center (CPC) update put El Niño emergence at 82% for May through July and at 96% for the heart of winter. The models have settled the direction. They have not settled the size, and on that, the same forecast is blunt: no strength category clears a 37% chance. This piece maps what the climate clock has already settled, what it has left open, and the five planting windows that close before its peak is visible.

Direction is settled

Start with what is no longer in doubt. The CPC’s mid-May discussion has El Niño forming within two months and persisting through the 2026-27 Northern Hemisphere winter at near-certain odds. The ocean is already primed: the weekly Niño-3.4 index sat at +0.4 °C in early May 2026 (NOAA), and the subsurface heat that feeds a developing event has risen for six consecutive months. A downwelling wave is loading the system. The European Centre for Medium-Range Weather Forecasts (ECMWF) and the United States’ own coupled model agree on direction. So does Brazil’s national met service, INMET, which now puts El Niño formation above 60% for the mid-year quarter and higher still in the second half.

When the models converge like this in one direction, the only question left worth pricing is how big.

Size is the open question

Here, the same forecast that settles direction refuses to settle scale. The CPC is explicit: there is substantial uncertainty in the peak strength, with no single strength category emerging as the most likely, and it adds that stronger events do not ensure stronger impacts; they only make certain impacts more likely. The agency that just called near-certain emergence will not name the most likely size.

Higher fertilizer prices may not be the biggest agricultural price threat this year; 2026 might produce a super El Niño weather pattern.

— Paul Donovan, Chief Economist, UBS

The model spread shows why. In ECMWF’s April multi-system ensemble, members for the autumn peak ran from barely above neutral to well past the super-El-Niño threshold; its own SEAS5 system spanned roughly +1.7 °C to +3.3 °C for September. The same models that agree El Niño is coming disagree, by a full category, on how strong it lands. Michelle L’Heureux, the CPC’s lead El Niño forecaster, ties the resolution to a process still in progress: a stronger El Niño is more likely “if changes in the atmosphere continue to sync with changes in the tropical Pacific Ocean this summer.” The coupling has not locked in. Until it does, the size is a live range.

Two honest complications. First, the upper tail is fattening: the odds of a very strong event peaking next winter have moved from roughly one in four to one in three in a single month. The modal outcome is unsettled, and the dangerous outcome is getting likelier — both are true. Second, the agencies do not even agree that the event has begun. Australia’s Bureau of Meteorology (BoM) still reads neutral because its threshold is wider than NOAA’s, and it requires the atmosphere to respond before it will declare an event. Same ocean, two verdicts. That gap is the argument for treating this as scenario analysis rather than a forecast.

It is also why the climate clock now outranks the others. Paul Donovan, chief economist at UBS, put the order plainly this spring: “higher fertilizer prices may not be the biggest agricultural price threat this year; 2026 might produce a super El Niño weather pattern,” and "In that case, drought and limited water supply might be more important than shortages of nitrogen.” The fertilizer clock has largely been printed. The climate clock is the one still moving.

WHERE THIS RESOLVES

On June 30th, we turn the five open questions into a full commodity-exposure map across the regions and crops in scope here.

Reserve a seat for the 30 June briefing →

Five windows, and only one genuine wildcard

Intensity matters because the regions exposed to it are not waiting for the winter peak to commit. Five planting and phenology windows resolve between now and September, four of them before a strong sea-surface-temperature signal would even be visible.

In India, the India Meteorological Department (IMD) has already cut its 2026 monsoon outlook to below normal, 90% of the Long Period Average (LPA), and pushed the onset back toward the start of June, citing the near-certain El Niño across the season and a counter-driver, the Indian Ocean Dipole, that has faded to neutral. The IMD’s June call is stark: “below-normal monthly rainfall is very likely over most parts of the country.” For the Kharif crop, where sowing timing is everything, a late and dry June is the hit. The last super El Niño, in 2015-16, ended at 86% of LPA and widespread drought.

In Brazil, El Niño splits the country, being wetter across the south and drier across the north and the Center-West frontier, but the exposure is deferred. The second-crop (safrinha) corn now in harvest was planted before this event could touch it; what is genuinely at risk is the 2026/27 crop. That decision opens with the September soy planting and runs straight into the event’s winter peak, when the dry side could pressure Center-West and Matopiba yields while excess rain complicates the southern harvest, a hit that would land in next year’s balance, not this one. In Australia, BoM leans the eastern winter-crop belt, New South Wales, Victoria, and South Australia, toward below-average rain through August, a 60–80% probability heading into grain fill. And across Southeast Asia, Indonesia’s met agency BMKG forecasts an early, prolonged, below-normal dry season for the palm belt. The cruelty there is the lag: the dose decision is made this year, and the yield miss prints in 2027.

Then there is the one that refuses to lean. The US Corn Belt is where El Niño’s summer teleconnection is weakest, and the CPC’s outlook reflects it: a slight dry tilt across the central US, much of the Plains in “equal chances,” no clean signal. This is the genuine wildcard; the market’s most-watched crop sits in the one window where intensity buys the least predictive power. Where four windows lean, the Corn Belt only widens.

What the last El Niños actually cost

The reason to size scenarios rather than call a number is written into the recent record. Three features recur, and all three argue against treating any single intensity outcome as fate. El Niño’s production hit is lagged; it usually prints in the marketing year after the ocean peaks. It frequently runs milder than the in-season fear. And it is amplified by policy at least as much as by weather.

We know how the dice are loaded.

— Weston Anderson, Famine Early Warning Systems Network (FEWS NET)

Palm is the cleanest analogue. The 2015-16 super event cut Malaysian palm output by roughly an eighth the following year, according to Malaysian Palm Oil Board (MPOB) data, and pushed Indonesia to its first palm export decline in eighteen years. Because the crop’s response runs six to twenty-two months behind the weather, a dry 2026 dry season books its miss in 2027 — the year buyers are not yet hedging.

The counter-case matters as much as the hit. The 2023/24 event ran milder than mid-season forecasts on rice — Indian production beat expectations and Delhi unwound its export curbs within a year — even as the same event drove drought disaster declarations across the Southern African maize belt. One event, two tails. The agencies that warned hardest on rice over-forecast the loss, and that miss is the intensity-uncertainty thesis in historical form. Weston Anderson of the Famine Early Warning Systems Network (FEWS NET) stated the established research view plainly: “we know how the dice are loaded.” Loaded dice still roll a range.

One crop runs the other way: global soybean yields tend to improve in El Niño years, an asymmetry we develop later in the series.

Price often moves on policy before it moves on yield. India’s 2023 ban on non-basmati rice exports, declared partly on El Niño fear, lifted the global benchmark by roughly a fifth, well past what the harvest shortfall alone warranted. The export-restriction cascade, which the Strait analysis flagged, is the channel to watch first; it front-runs the agronomy.

What will the next ninety days decide

The climate clock has answered the cheap half of its question and left the expensive half open. El Niño is coming; its size is unresolved; and the windows that translate size into supply are closing on a calendar that does not wait for the winter peak. The right posture is a set of scenarios sized to a strength range that still runs from moderate to super, with a watch on which of the five windows breaks first.

We will resolve those scenarios into a full commodity exposure map — what the five questions resolve to, live on June 30th. Risk and procurement teams sizing 2027 positions should join us there: the preview names the questions; the webinar provides the answers.

Frequently Asked Questions

Will El Niño form in 2026, and how certain is it?

Emergence is near-certain. The NOAA Climate Prediction Center’s mid-May 2026 discussion put El Niño formation at 82% for May through July and at 96% for the December 2026–February 2027 winter, and the European Centre for Medium-Range Weather Forecasts (ECMWF) and Brazil’s INMET agree on direction. The open question is not whether the event arrives but how strong it becomes.

How strong will the 2026–27 El Niño be?

That is unresolved. The same NOAA CPC forecast that calls near-certain emergence assigns no single strength category more than a 37% chance, and the odds of a very strong (“super”) event peaking next winter have risen from roughly one in four to one in three in a single month. The intensity, not the direction, is what prices into 2027.

Why do forecasters agree El Niño is coming but not on its strength?

Because the ocean has primed but the atmosphere has not yet locked in. ECMWF’s April multi-system ensemble spanned roughly +1.7 °C to +3.3 °C for the September Niño-3.4 anomaly — a full El Niño category of spread. CPC forecaster Michelle L’Heureux ties the resolution to whether atmosphere–ocean coupling syncs over the summer.

Which crops and regions are most exposed before the winter peak?

Five planting and phenology windows resolve between now and September. India’s monsoon outlook has already been cut to 90% of the Long Period Average; Brazil’s exposure falls on the 2026/27 crop now being planted, not the safrinha in harvest; Australia’s eastern winter-crop belt carries a 60–80% chance of below-average rain; and Indonesia’s BMKG forecasts an early, prolonged dry season for the palm belt. The US Corn Belt is the wildcard, where the El Niño teleconnection is weakest.

When does an El Niño actually hit crop supply?

Typically with a lag. The production hit usually prints in the marketing year after the ocean peaks: the 2015–16 super event cut Malaysian palm output by roughly an eighth the following year (MPOB), and because palm’s response runs six to twenty-two months behind the weather, a dry 2026 dry season books its miss in 2027. El Niño impacts are also frequently milder than the in-season fear.

Why does Australia’s BoM read neutral while NOAA calls El Niño near-certain?

Different thresholds. Australia’s Bureau of Meteorology (BoM) uses a wider sea-surface-temperature threshold than NOAA and requires the atmosphere to respond before it declares an event, so it still reads neutral on the same ocean NOAA reads as near-certain. The gap is a reason to treat the outlook as scenario analysis rather than a single forecast.

What should procurement and risk teams do with an unresolved intensity outlook?

Size scenarios rather than call a number. Because the strength range still runs from moderate to super and policy responses (such as India’s 2023 rice export ban, which lifted the benchmark by roughly a fifth) often move prices before yields do, the durable posture is a set of scenarios across that range, watching which of the five regional windows breaks first.

JOIN THE WEBINAR

Risk and procurement teams sizing 2027 positions get the full commodity-exposure map across all five windows — the preview names the questions, the briefing prices the answers.