Nobody in Ghanaian cocoa sells at the price on the screen. A state board fixes the farmgate rate, announces it by circular, and revises it when politics requires. This pillar sets out which price is which, who moves each one, who buys the beans, and why half of Ghana’s grinding capacity stands idle.
The price in the headlines is not the price anyone in Ghana transacts at. The world reference price averaged US$7.80 a kilogram in 2025, against US$7.33 in 20241, and is forecast at US$3.80 a kilogram for 2026, a fall of 51.3 per cent, recovering to US$4.20 in 2027; it fell 32 per cent in the first quarter of 2026, to its lowest level since the third quarter of 2023. COCOBOD then fixes a gross free-on-board value for the season, put at US$7,200 a tonne, a season target rather than a realised sale price2, and pays the farmer a share of that rather than of the screen. A futures move reaches a Ghanaian farm only when a circular carries it there.
The farmer’s share of the world price is a matter of policy rather than of contract, and in practice the government targets about 70 per cent of the free-on-board price, and the producer price is on occasion pegged significantly below or above it3. And the percentage depends on whose exchange rate is used. USDA put the August 2025 price at 68 per cent of that target4, COCOBOD at 70 per cent of it, up from 63.9 per cent the season before5, on the same cedi figure. Anyone modelling a farmgate price off an FOB percentage is really modelling a cedi assumption.
The living-income top-up most entrants ask about is not payable. Fairtrade’s reference price for Ghanaian cocoa is US$3,408 a tonne for the season from 1 October 2025, and the differential against it is not applicable, zero, because the announced farmgate price of US$4,662 a tonne already exceeds the reference price. It is paid only by buyers who have voluntarily committed to it, and only where the regulated farmgate sits below the reference, which Ghana’s does not. Put nothing in the model for it.6
The only price you transact at| Price layer | Level as last stated | Who sets it | What it means for an entrant |
|---|---|---|---|
| World reference price | US$7.80/kg average 2025; US$3.80/kg forecast 2026 | World futures markets; annual average published by the World Bank | Sets the ceiling COCOBOD prices against. You cannot sell at it. |
| Gross FOB, 2025/26 | US$7,200/tonne (a target, not a realised sale) | COCOBOD | The base every downstream share is cut from. No realised 2025/26 FOB is held here; USDA gives US$2,600/MT for 2023/24, COCOBOD US$4,850/MT for 2024/25. |
| Farmgate, main crop | GH¢58,000/tonne since 3 October 2025 | COCOBOD circular | The only price a Ghanaian transaction uses. Rendered at US$4,622 by USDA and US$4,662 by Fairtrade: two rates, one cedi figure. |
| Farmgate, light crop | GH¢41,392/tonne from 18 June 2026 | COCOBOD circular | About 71 per cent of the main-crop rate, on 20 to 30 per cent of the year's volume. |
| Living Income Reference Price | US$3,408/tonne; differential n/a | Fairtrade International | No top-up payable this season: the regulated price is already above the reference. |
Sources: World Bank Group, Commodity Markets Outlook (April 2026); COCOBOD producer price release (4 August 2025); USDA FAS GAIN Semi-Annual (December 2025); Ghana News Agency (16 June 2026); Veldhuyzen van Zanten, Fairtrade International (8 October 2025). The FOB figure is a season target, not a realised average sale price.
Only one layer in this table is a price you transact at, the farmgate, and it is announced rather than discovered. What an entrant chooses is a position against it: which discount you buy at, and which currency you carry the risk in.
On 4 August 2025, effective the 7th, the main crop opened at GH¢3,228.75 for a 64 kg bag, or GH¢51,660 a tonne. On 2 October, effective the 3rd, it was revised up to GH¢3,625 a bag, or GH¢58,000 a tonne. The trigger was not a market move: farmer protest after Côte d’Ivoire set its own 2025/26 farmgate at XOF 2,800 a kilogram, about US$5.00, on 1 October. In dollars per bag that revision was a cut, from about US$306 to US$288.89, because the cedi fell further between the two dates than the price rose.7
On 16 June 2026 COCOBOD issued the light-crop circular, effective for purchases from 18 June, fixing the price at GH¢2,587 a bag, or GH¢41,392 a tonne, for Grade I and Grade II alike, ex-scale at every designated buying centre8, with no stated rationale and no comparison to the previous light crop. That is roughly 71 per cent of the main-crop rate on real volume, because the light crop carries 20 to 30 per cent of annual production9. Light-crop beans are smaller and sell at a 12.5 per cent discount on the international market, so the discount is carried twice, once at each end.
Two forward signals sit outside the circulars and both rest on one weak source, a contracts-for-difference broker’s page. It records cocoa spot at US$5,700 a tonne on 13 July 2026, up sharply from a March 2026 low near US$2,863 and below a 24-week high of US$6,520, far above the World Bank’s forecast, and a Ghana and Côte d’Ivoire joint declaration signed on 16 June 2026 to align farmgate pricing and harmonise crop calendars from the 2026/27 season. Nothing else here carries either. Watch them; do not plan on them.10
Announced producer prices, not transacted averages. The two dollar figures use the two different exchange rates USDA applies to the two announcement dates (GH¢10.541 and GH¢12.548), so the fall between them is a currency movement, not a price cut. No source gives a dollar equivalent or an FOB percentage for the June 2026 price. Cedi per-tonne figures run on a 16-bag basis; the announced unit is the 64 kg bag.
Cocoa is Ghana’s second largest source of foreign exchange in the first half of calendar 2025, on US$2.47 billion of exports, between gold at US$8.7 billion and petroleum at US$1.36 billion. On the wider marketing-year measure, export earnings from beans and cocoa products surpassed US$3.37 billion over October 2024 to June 2025, against US$1.54 billion for the whole of the previous marketing year. The buyer list behind that money is short. Between October 2024 and June 2025 the Netherlands alone took 106,025 tonnes, the United States 45,961 tonnes, Belgium 31,784, Malaysia 30,572 and Japan 29,137, out of 362,521 tonnes of beans shipped. Five destinations, two thirds of the volume.11
The concentration carries a direction, and that direction is what creates the exposure. USDA reports 292,755 tonnes of Ghana’s 512,515-tonne reported global export volume for 2024 going to the European Union and calls this close to 75 per cent; the two figures give 57 per cent. Deforestation-regulation compliance is therefore not a certification upgrade but a condition of market access for the largest single destination bloc, more than half the crop; Pillar 3 carries the deadlines and the cost. The American exposure is already priced in: a 15 per cent country-specific reciprocal tariff on Ghana, announced 31 July 2025 and applied from 7 August, and AGOA, whose main benefit to cocoa was duty-free access for processed products, expired on 30 September 2025. Raw beans generally entered the United States duty-free anyway, so both changes fall hardest on the processed products Ghana is trying to sell more of.12
Cocoa beans only, nine months to June 2025, not a full marketing year and not including paste, butter or powder. The five named destinations are the only ones USDA lists; the balance bar is the remainder to the 362,521 tonne period total and is not itself a reported figure. The 75 per cent EU share is a separate measure: all products, calendar 2024.
Installed domestic grinding capacity is 504,780 tonnes, and processors run at 50 per cent of it or below, on insufficient and inconsistent bean supply. Deliveries to domestic grinders were 178,665 tonnes in 2023/24 and 235,472 in 2024/25, the latter in a season cut two months short, and grindings are forecast at 300,000 tonnes in 2025/26, up 36 per cent on the preceding year’s grindings estimate of 220,000 tonnes. Half of Ghana’s grinding capacity standing idle is not a machinery problem but an allocation problem, and the allocator is COCOBOD.13
The entire incentive to grind in Ghana is one administered number: beans delivered to domestic grinders go at a 20 per cent discount to the main-crop bean price. No source in this pillar’s evidence base describes a statutory grinding quota or a local-processing mandate, and this playbook will not invent one. The return on a processing investment therefore turns on a bean allocation, which is a relationship and not a contract. Buy utilisation in a plant that already has deliveries.14
Half the plant, none of the beans
The farmgate price is fixed by circular in cedis and revised on political triggers, so it tracks neither the futures market nor the dollar. Between the August and October 2025 circulars the cedi price rose while the dollar price per bag fell.
Producer price GH¢51,660 per tonne from 7 August 2025 and GH¢58,000 from 3 October, rendered at about US$306 and US$288.89 a bag. World reference prices are forecast to fall 51.3 per cent in 2026.
Anyone carrying dollar costs or dollar debt against cedi cocoa revenue; investors underwriting an export margin off a futures curve.
Model in cedis and stress the dollar separately. Assume the farmgate lags the world price in both directions, and treat the 70 per cent FOB rule as a target already missed both ways, not as a floor.
Ghana's grinding capacity runs at half utilisation because beans are allocated rather than bought on an open market. A plant without a standing allocation is a fixed cost with no throughput, and no statutory quota exists to compel supply.
Installed capacity 504,780 tonnes against processors operating at 50 per cent or below on insufficient and inconsistent supply, with deliveries to domestic grinders of 178,665 tonnes in 2023/24.
Entrepreneurs and investors weighing a grinding, butter or liquor plant; lenders underwriting one.
Contract the bean allocation before the equipment. Test the 20 per cent grinder discount against three seasons of delivered volume, never against installed capacity. An under-used plant with a delivery record beats a new one without.
Trade the administered discounts, not the world price. Decide first which side of the 12.5 per cent mid-crop discount and the 20 per cent grinder discount you are trading on, because those two administered numbers are where the margin sits, not the world price. The trap is committing to a dollar-priced input or loan against cedi cocoa revenue: the October 2025 revision raised the cedi price and cut the dollar price per bag in the same circular.
Buy a delivery record, not installed capacity. Contract the bean allocation before the equipment. Test the 20 per cent grinder discount against three seasons of delivered volume, never against installed capacity. An under-used plant with a delivery record beats a new one without.
Model in cedis, and price the currency mismatch as a service. The farmgate is fixed by circular and tracks neither the futures market nor the dollar. Model in cedis and stress the dollar separately, and treat the 70 per cent FOB rule as a target already missed both ways, not as a floor.
Four prices sit between the world market and the farm gate, and only the COCOBOD farmgate rate is one you can transact at. It is announced by circular, in cedis, and revised on political triggers.
The 2025/26 sequence ran GH¢51,660 a tonne from 7 August 2025, GH¢58,000 from 3 October, and GH¢41,392 for the light crop from 18 June 2026. The October rise was a fall in dollars per bag.
No living-income top-up is payable this season. Fairtrade's reference price is US$3,408 a tonne and the regulated farmgate already exceeds it, so the differential is zero.
USDA reports 292,755 of 512,515 tonnes of Ghana's 2024 global exports going to the European Union and calls this close to 75 per cent; the two figures give 57 per cent, still the largest single bloc. Five destinations took two thirds of bean volume in the nine months to June 2025. EU compliance is market access, not a premium.
Installed grinding capacity is 504,780 tonnes and runs at half of it or less on bean supply. The constraint on processing is allocation, not machinery, and there is no statutory quota to force it.
Where this connects. Pillar 3 owns the deforestation regulation behind the EU market share priced here. Pillar 5 turns the farmgate price into a margin against what production costs. Pillar 7 goes behind the destination table to the buyers themselves.