Cocoa: a real shortage, then a very fast unwind

The International Cocoa Organization's daily-price average for 2022 was US$2,368 a tonne. Then West Africa had two bad seasons in a row — and Africa, overwhelmingly Côte d'Ivoire and Ghana, grew 73.6% of the world's cocoa in ICCO's 2022/23 estimate. Heavy rain from mid-2023 spread black pod disease through Côte d'Ivoire and Ghana, and an unusually strong, dry Harmattan wind and record heat followed in early 2024. Côte d'Ivoire's port arrivals for 2023/24 ran about 25% below the year before; Ghana delivered roughly 432,000 tonnes against an 800,000-tonne target. By the time ICCO closed its books on that season, world output had fallen 12.9%, the market was short 494,000 tonnes, and stocks had dropped to 26.4% of a year's grindings, the lowest ratio in ICCO's series.

The unwind was as fast as the rise. High prices cut demand — Barry Callebaut, the world's largest cocoa processor and chocolate supplier to food companies, reported a 6.8% volume fall for its 2024/25 year — and the 2024/25 crop recovered 8.5% to 4.733 million tonnes, turning the deficit into a 37,000-tonne surplus. New York futures fell below US$4,000 in January 2026 and bottomed near US$2,850 in the spring. They have since climbed back towards US$6,000 on renewed black pod reports and El Niño risk for the 2026/27 crop. For the current season there is no official balance to check that against: ICCO's August 2026 bulletin says its Secretariat has "temporarily withheld" 2025/26 production and grindings data.

How much of cocoa's spike was climate change?

Three pieces of evidence get cited, and each says something narrower than the headlines built on it.

• World Weather Attribution (March 2024) studied the 11–15 February 2024 heatwave in coastal southern West Africa and found human-caused warming made its humid-heat index about 4°C higher and the event roughly ten times more likely. It is a heat-and-health study; it did not assess cocoa yields or prices.
• Climate Central (February 2025) looked at 44 cocoa-growing regions and found climate change added at least three weeks a year of days above 32°C — a heat-stress threshold for cocoa trees — to the October–March main-crop season in Côte d'Ivoire and Ghana over the past decade, and about six extra weeks in 2024 in 71% of the areas studied. It says heat "can contribute to" lower yields and quality alongside rain and disease, and does not put a number on the price effect. It is an analysis by a research non-profit, not a peer-reviewed paper.
• Kotz and co-authors (Environmental Research Letters, July 2025), a team including European Central Bank economists, catalogued 16 food-price spikes that followed weather extremes in 2022–24. Its cocoa entry notes unprecedented monthly heat across most of Ghana and Côte d'Ivoire in February 2024, on top of earlier drought, followed by world cocoa prices up about 300% year on year by April 2024. It is a perspective article that shows the two happened together; it does not estimate how much of the rise the weather caused.

The honest reading is that climate change made the weather that triggered the shortfall worse, and that nobody has yet measured how much of a US$10,000 price move that accounts for. The shortfall landed on a crop that was already fragile for reasons that have nothing to do with the atmosphere. Ghana's cocoa board, COCOBOD, has reported about 590,000 hectares infected with cocoa swollen shoot virus, a disease with no cure except cutting out the tree. Illegal gold mining has taken roughly 19,000 hectares of cocoa land in Ghana. Two decades of low, state-fixed farmgate prices left orchards old and under-fertilised, and pushed an estimated 160,000 tonnes of Ghana's 2023/24 crop to be smuggled across borders where prices were higher. On the exchange, margin calls drove hedgers out and liquidity hit record lows in the spring of 2024, which amplified a supply problem measured in hundreds of thousands of tonnes into outsized daily price swings.

On the way down, farmers took the cut; shoppers mostly didn't

West African governments raised farmgate prices late and cut them fast. Côte d'Ivoire's coffee-cocoa council set a record 2,800 CFA francs a kilo for the 2025/26 main crop, then cut the mid-crop price 57% to 1,200 in March 2026 and kept it there for 2026/27. Ghana raised its producer price to GH¢58,000 a tonne in October 2025, cut it to GH¢41,392 in February 2026, and has set GH¢42,400 for 2026/27 on the assumption of a US$2,650 world price.

Côte d'Ivoire: main-crop price; the 2025/26 cut to 1,200 applied to the March 2026 mid-crop. Ghana: arrows show in-season revisions, April 2024 and February 2026.

Chocolate makers moved the other way on a lag. Hershey announced a "low double-digit" price rise in July 2025; Lindt raised prices 15.8% in the first half of 2025. Because manufacturers buy cocoa about a year forward, Mondelez has said it does not expect meaningful relief until 2027. Through mid-2026, US press reports found shelf prices had not come back down despite cocoa falling about 70% from its peak, with makers turning to promotions and new pack sizes rather than list-price cuts. The price spike reached consumers in months; the relief, so far, has reached them mainly as discounts.

Vanilla: two weather spikes, and a crash that wasn't the weather

Vanilla is the more extreme case of a crop grown in one place. Madagascar supplied 54% of world vanilla exports by value in 2023, and most of the cured Bourbon vanilla the flavour trade actually prizes comes from one region on its north-east coast, SAVA. (The popular figure of "80% of world vanilla" is not what UN or FAO data show; it appears to describe Madagascar's share of cured Bourbon beans.) Natural vanilla also supplies a sliver of the vanilla flavour people eat — well under 1% by most estimates, the rest being synthetic or bio-based vanillin — which is why its price can swing so far without anyone noticing at the supermarket.

Two of vanilla's price spikes do trace to cyclones that struck SAVA directly. Cyclone Hudah, in April 2000, destroyed about 4,355 of the region's 25,500 hectares of vanilla, according to an FAO/WFP mission; trade prices then climbed to around US$400–500 a kilo in 2003–04 before collapsing to about US$30 by 2005. Cyclone Enawo, in March 2017, landed on the same coast after Nestlé and other food companies had begun switching from vanillin back to natural vanilla in 2015, and prices reached about US$600 a kilo that summer.

The collapse since 2019 is a different story. After the Enawo peak, high prices pulled in new planting and pushed farmers to pick early to beat theft. Madagascar's government tried to hold prices up with a minimum export price — US$350 a kilo in February 2020, lowered to US$250 that July — which buyers resisted; the IMF recorded vanilla exports collapsing by two-thirds after enforcement was tightened in late 2022, and unsold stocks built up. When the floor was scrapped in May 2023, the backlog came out at once: UN Comtrade data show Madagascar's 2024 export volume at 3.2 times 2023's while the value fell 15%. By the fourth quarter of 2025 the central bank's average export price was US$38 a kilo, and a government-backed Vanilla Alliance is now trying to buy 600 tonnes off the market. Cyclone Gamane did hit SAVA in March 2024 and damaged crops, but prices kept falling through it. The two big cyclones of 2022 and 2023 that are sometimes linked to vanilla, Batsirai and Freddy, made landfall far to the south of the vanilla coast.

The climate evidence for Madagascar is also thinner than it looks. The best-known attribution study, World Weather Attribution's December 2021 analysis of the southern Madagascar drought, found the drought's likelihood had "not significantly increased" due to human-caused climate change, with poverty, rain-fed farming and the pandemic doing more of the damage. A 2024 paper in npj Climate and Atmospheric Science by Rigden and co-authors disagreed, finding human-driven soil-moisture trends made the 2017–22 drought significantly more likely. Neither study covers SAVA, and this piece found no formal attribution study of any cyclone that hit the vanilla region. What does exist for SAVA is a 2025 survey of about 500 farmers by Duke University and the University of Antananarivo: nearly all report changing temperatures and rainfall, which is evidence of what farmers experience rather than a trend in weather records.

India grows a little of both, and has done this before

Cocoa. India grows cocoa mainly as an intercrop under coconut and arecanut palms, chiefly in Andhra Pradesh and Kerala, with smaller areas in Karnataka and Tamil Nadu. Trade and aggregator sources put national output at roughly 27,000–30,000 tonnes a year from about 1.1 lakh hectares, with Andhra Pradesh the largest producer at around 40%; the official state-wise table from the Directorate of Cashewnut and Cocoa Development was tabled in a Rajya Sabha answer in December 2024 and should be the reference for exact numbers. That is a fraction of what India's chocolate industry uses, and the gap is imported. UN Comtrade data show India imported 20,657 tonnes of cocoa beans worth US$62.3 million in 2023 and 28,774 tonnes worth US$186.8 million in 2024 — volume up 39%, the bill tripled, and the average price per tonne more than doubled to about US$6,490. Most of those beans came from DR Congo, Ecuador and the Dominican Republic rather than West Africa, which did not help: the price is set in London and New York, so an Indian chocolate maker pays the West African shortage price no matter where its beans come from.

For Indian growers the spike was a windfall while it lasted. Dry beans that had fetched ₹150–250 a kilo in Kerala reached about ₹830 in Idukki in April 2024 and around ₹1,000 by May, before easing to about ₹550 by September. By 2025 prices had fallen far enough that Andhra Pradesh announced a guaranteed purchase price of ₹500 a kilo through Mondelez. The state's horticulture department has reportedly set a goal of adding about 5,000 hectares of cocoa a year.

Vanilla. India has already run the experiment of planting into a price peak. When Cyclone Hudah's shortage pushed prices up, Kerala and Karnataka farmers planted vanilla on a large scale: green beans fetched about ₹3,500–4,000 a kilo in 2003–04, Karnataka's vanilla area went from 265 hectares in 2000–01 to more than 2,000 in 2004–05, and the national area peaked at 5,815 hectares in 2005–06; in 2006–07 India produced 233 tonnes of cured beans. Then world prices fell to around US$30 a kilo, green beans in India fell to about ₹80, and by 2012 the area had shrunk to under 200 hectares. India produced 44 tonnes of vanilla in 2024–25, down from 140 tonnes in 2018–19, and it is now a net buyer: it imported 57.6 tonnes worth US$4.33 million in 2023, about half of it from Madagascar, and exported 14.8 tonnes worth US$1.73 million in 2024.

What would fix this

• The Directorate of Cashewnut and Cocoa Development should publish state-wise cocoa area, production and a monthly farmgate price series on its own website, updated each season. Today the state table surfaces mainly through parliamentary answers and farm prices only through press reports, which makes it impossible for a grower to judge whether new planting makes sense.
• Andhra Pradesh's Horticulture Department should tie its annual cocoa area-expansion target to contracted offtake volumes and publish both, so that new planting tracks demand buyers have committed to (such as the Mondelez purchase arrangement) rather than the 2024 price peak — the mistake India made with vanilla in 2003–05.
• The Spices Board should restore vanilla to its published area, production and export tables and update its vanilla crop page, which still describes an area of about 1,000 hectares, so farmers can see that the crop is now a 44-tonne net import rather than the export opportunity it was sold as two decades ago.
• ICAR's Central Plantation Crops Research Institute and Kerala Agricultural University should publish multi-season yield and bean-quality results for the cocoa varieties they distribute, measured under heat and drought stress, since heat and disease drove West Africa's shortfall and India's expansion is aimed at the same buyers.

These are proposals from this blog, not recommendations any agency has adopted.

Sources and caveats

About this article: Researched, written and edited by Umashankar Triplicane Dwarakanathan, with AI research assistance; every figure is meant to trace to the primary source cited. See the Editorial Policy for how sourcing, AI use and corrections work.