Cocoa Market Review May 2026

OVERVIEW

For all the bearishness of the current fundamentals, a record surplus, strong stock to grind ratio, and physical differentials at multi year lows, cocoa spent May rallying, largely because the futures market has stopped looking at 2025/26 and has turned its attention to the 2026/27 season.

What is now driving the market is the strength and timing of an El Niño event that is now very likely to emerge, but whose severity remains deeply uncertain. It is that uncertainty, rather than current fundamentals, that is driving volatility and keeping the market’s view of risk skewed to the upside.

FLAT PRICE & POSITIONING

May saw a sharp increase in volatility as the market’s attention turned towards the development of the 2026/27 crop. New York July traded from a low of $3,487 to a high of $4,777 before settling the month at $3,923, up $354. London July closed at £2,975, up £307, having reached a high of £3,549. The move was accompanied by significant short covering, with the market breaking through a number of technical levels including the support that had held in January and early February prior to the aggressive origin selling seen earlier in the year. The growing probability of an El Niño event likely also played a part in the rally.

The combined managed money net short position across both markets reduced by 6,847 lots over the month, from 45,016 on 5 May to 38,169 on 26 May. The bulk of the covering was front loaded into the first week, with the net short falling sharply from 45,016 to 33,956 as the rally took hold, before managed money began rebuilding shorts into the back half of the month, adding 3,459 lots in the final week even as price held up. This suggests bearish conviction has not fully capitulated and that the short covering potential remains very much intact should a further bullish catalyst emerge.

Origin sold into the short covering rally, although we do not believe the CCC and CMC were overly active. Ivory Coast forward sales for 2026/27 stand at approximately 1.1 million tonnes, with Ghana at 200,000 to 250,000 tonnes, Ivory Coast continuing to sell at levels over ICE London Cocoa. There was also a fair amount of residual hedging from exporters who were previously unhedged. This leaves the CCC and CMC with an estimated 500,000 tonnes still to sell by September. Their relative absence at these price levels continues to look like a missed opportunity. Having been aggressive sellers since February at levels below where the market currently sits, the urgency has clearly diminished, but the need to sell has not gone away.

Industry was largely inactive during the month, having added good cover prior to th e short covering rally. However, towards the back end of the month it appears they did participate at higher levels, with the market trading in a $3,700 to $3,900 range basis July 2026 New York and £2,800 July 2026 London. We had anticipated the market would test lower levels before finding meaningful industry support. In the event, the combination of El Niño uncertainty and poor early pod count reports appears to have brought industry in earlier and at higher levels than the fundamental surplus picture alone would have warranted, and we would expect them to continue paying up from the levels at which they have been buying since February to April.

Our overall view is that downside is likely limited over the next couple of months, given the combination of industry support and a still large speculative short. Beyond that, however, the conditionality is important. The underlying fundamentals for 2025/26 are extremely bearish, and the high stock to grind ratio provides a cushion such that even a small to moderate deficit for 2026/27 would leave the market comfortable. The bullish case therefore rests almost entirely on an El Niño materialising and having a visible effect on production. Without a major production shock, we would expect the surplus to reassert itself and the market to come off from these levels.

2025/26

Market estimates for the 2025/26 surplus span a wide range, from around 320,000 to just under 600,000 tonnes. Drawing on the range of market estimates, we would place the surplus at around 450,000 to 500,000 tonnes. The CCC raised its crop guidance to 2.1 to 2.2 million tonnes, from 1.8 to 1.9 million tonnes, though the market has long been working with a larger crop than the previous official figure.

The CCC announced arrivals of over 1.7 million tonnes by 11 May, around 130,000 tonnes above the Reuters figure. Arrivals since 27 April have totalled 126,000 tonnes, up from 98,000 tonnes in the same period last year, and we expect this trend to continue through the bulk of the mid-crop. Soil moisture is reported to be at its highest for this time of year and weather conditions have been conducive to mid-crop development. The mid-crop arrival period is where we expect the surplus relative to last year to become most visible.

Ecuador bean and product exports for April were again strong at 38,053 tonnes against 36,656 tonnes in April 2025, keeping the country on course for another record crop. Ecuador continues to trade at a discount of around $250 to $300 under New York at origin, tendering competitively in both markets. Nigeria bean exports for April were down 20% year on year at 14,921 tonnes, following a sharper 35% decline in March.

Ivory Coast grinding for April came in at 46,578 tonnes, up 22% on April 2025, though this partly reflects a weak comparative; year to date grindings stand at 381,349 tonnes, down 2.1% on the same period last year. Of note is a shift in grinding dynamics, with cheap origin differentials and improved crop availability redirecting some grinding activity back to Ivory Coast from Europe.

2026/27

Early pod count reports for the 2026/27 crop setting in West Africa have been poor, although it is far too early to draw meaningful conclusions, as reports were similarly disappointing at this point last year. The next round of pod counts in June will provide a clearer picture.

The market remains very wary of the likely emerging El Niño over the coming months. The latest NOAA update estimates an 82% chance of an El Niño emerging in the May to July window, rising to 96% of an event extending towards December 2026 to February 2027, with up to a 67% chance of a strong to very strong event developing later in 2026. Historically, the production impact scales with the strength of the event: weaker El Niños tend to have a relatively modest effect, while a strong to very strong event can significantly reduce output.

The forecast through to September looks relatively benign in West Africa, with neutral to slightly positive probabilities for better rainfall, meaning the early development of the 2026/27 crop should be largely unaffected for now; the greater risk is to the West African crop later in the year. In Ecuador the expected El Niño footprint is already emerging, with a 70 to 100% chance of above normal precipitation in coastal regions, and a similar 70 to 100% chance of drier than normal conditions in Indonesia. A strong rise in Niño 1+2 sea surface temperatures would raise the risk of flooding in Ecuador’s growing regions, a risk that, given the country’s rapid emergence as a major origin, may be underappreciated relative to the market’s focus on West Africa.

A further compounding risk for 2026/27 yields is the sharp rise in fertiliser costs following the closure of the Strait of Hormuz; should an adverse weather event materialise simultaneously, the impact on production could be amplified, with some farmers likely to reduce or skip fertiliser application altogether due to the increase in price.

On demand, while we remain some way from a return to 2023/24 levels, there are signs of improvement for 2026/27. Given the relatively low prices of 2026 compared with the prior two years, the market anticipates an improvement in grind of around 5%. Hershey has made public its plans to increase the cocoa content in its chocolate alternatives, while Mondelez has trimmed prices on some European chocolate and reported rising sales volumes. That said, some of the market remain sceptical of the improvement in grind and the durability of demand from reformulated products; with the risk of El Niño ahead, manufacturers may be slower to move back from reformulated products than anticipated.

MARKET STRUCTURE

London structure firmed through the month. The July/September spread (NU) moved from a contango of -12 to a backwardation of +21, having reached a high of +25. This strength is somewhat artificial given that only around 3,520 tonnes are expiring in July, but as demonstrated when the K spread approached a premium and attracted over 25,000 tonnes of grading, cheap origin differentials and a substantial surplus mean any sustained backwardation is likely to attract significant selling. The September/December spread (UZ) closed at -48, having weakened through the month; the September contract has around 22,000 tonnes expiring, with Nigeria the dominant origin, which could keep downward pressure on UZ. Further forward, December/March (ZH) was the weakest spread on the curve, trading to a low of -56 on the final day of the month before settling at -53, down £32 on the month, and we believe ZH could remain very weak as the last expiry before EUDR comes into force. Holders of non compliant cocoa are incentivised to move it through the December 2026 contract to avoid the £100 per tonne discount that will apply to non compliant cocoa delivered against 2027 contracts, which could weigh on December 2026. Looking ahead to 2027, ICE has increased delivery limits across those positions (3,750 lots for March 2027, 5,000 lots for May 2027, and 7,500 lots for the remaining positions), a somewhat counterintuitive move given that the amount of compliant cocoa eligible for delivery is likely to shrink as the industry comes to terms with EUDR’s traceability requirements. This raises the prospect of tightness in the 2027 positions should compliant cocoa prove scarce. That said, the risk should not be overstated: meaningful volumes of non compliant cocoa may still be deliverable in 2027, particularly if the Z/H switch offsets the £100 discount, and the timing of EUDR itself is not yet certain.

New York structure remained weak across the board. The July/September spread (NU) closed at -76, September/December (UZ) at -102, around full carry, and December/March (ZH) at -77, the curve in steep contango throughout. 1,912 lots are available to July, with 1,131 lots originating from Ecuador. Ecuador continues to trade at under tender parity at origin, at $250 to $300 under New York, which should continue to suppress the forward curve. Warehouse stocks in the US rose to 184,867 tonnes by the end of May, up around 41,884 tonnes year on year; while this is a notable increase and consistent with the surplus, absolute stock levels remain well below the average of the past decade.

THE ARBITRAGE

New York trades at a premium to London across most of the curve, with the exception of July, which is at a discount. The weakening of the arbitrage over the month is best explained by where the short covering was concentrated. London managed money covered aggressively, reducing its gross short by close to 15,700 lots and more than halving its net short from -33,823 to -20,021 over the month. New York managed money, by contrast, was largely static, with its net short little changed at around -18,000 lots. With the bulk of the speculative buying flowing through London, London structure firmed, a move reinforced by the small volume expiring in July London; New York, lacking that same support and with decent availability against July, remained under pressure.

PHYSICAL MARKET & DIFFERENTIALS

Cocoa bean differentials in the second hand market weakened materially over the month in response to the futures rally and the aggressive forward selling from Ivory Coast and Ghana since February. Ivory Coast nearby shipment differentials are at their lowest level since December 2017. Ivory Coast has sold close to 1 million tonnes of new crop, with Ghana at 200,000 to 250,000 tonnes for the same positions, and 2026/27 differentials, particularly West African, are largely unchanged on the month.

Product values are also slightly weaker, particulary butter given the higher market. Generally industry are well covered for both beans, and particularly products (butter and liquor) both in ratio and outright terms. Cocoa butter ratios, which traded down to 10 year lows in February, have since recovered somewhat and appear to have found a floor for the time being, a tentatively constructive signal for the demand side.