Value Chain Analysis · Cocoa · Ghana Agribusiness Playbook

Value Chain Analysis · Pillar 06

Ghana grows some of the best cocoa in the world and keeps almost none of what that cocoa is finally worth. The chain from a Western Region farm to a European shelf has nine links, and Ghana’s involvement ends at the third or fourth, when the beans leave the port in sacks. This pillar maps the nine links, prices the two ends where value concentrates, and shows that the one grinding step the country built stands half idle and, in the state’s own case, runs at a loss.

The chain that carries a bean from a farm in the Western Region to a chocolate bar on a European shelf has nine links, and the country’s involvement ends at the third or fourth of them, when the beans leave the port in sacks. Everything that makes cocoa valuable, the grinding, the blending, the branding and the retail markup, happens after Ghana has already been paid. This pillar maps the nine links, prices the two ends where the value actually concentrates, and shows that the one place Ghana has built a domestic value-addition step, the grinding factories, stands half idle and, in the state’s own case, runs at a loss.

Ghanaian farm workers carrying and stacking plain burlap sacks of dried cocoa beans into a tall pile in an open dirt yard, forested hills in the background, documentary photograph

Nine links, three paid to Ghana

Nine links carry a bean from this yard to a European shelf; Ghana is paid at about the third of them.

9 links

farm to shelf; Ghana's involvement ends at the third or fourth

~18%

the EU consumer-chocolate value Ghana retains

5-6%

the farm-gate share of a finished chocolate bar

A map of Ghana showing the share of each region's land under cocoa, concentrated in the south and west, with export ports and market hubs marked — Figure 9 Where Ghana’s cocoa actually is, and where its value chain starts.

What this shows

Cocoa is a southern crop, and in the west it is close to the only crop. The same concentration that makes the chain efficient also makes it exposed: one disease front or one European rule lands on the same few regions at once.

How to read this chain, and what the evidence can and cannot carry

The important feature of Ghana’s cocoa chain is not that any single actor is cheating the farmer. The producer price is administered, set once a season by a government committee, so the farm-gate transaction is the least negotiated link in the whole chain. The value leaks somewhere the farmer never sees, in the gap between what a tonne of raw beans earns and what the same tonne earns once it has been ground, moulded and branded abroad. Two held figures fix the size of that gap. Ghana retains about 18 per cent of the total value generated by the sale of consumer chocolate in European markets¹, and the farm-gate price is only about 5 to 6 per cent of the cost of a finished bar.²

The chain below is drawn link by link, and each link is described as structure rather than priced, because the domestic margins between farm-gate and export are fixed administratively and are not the place the money is made or lost. The two links that carry real, held financial evidence are the grinding link, where the state processor’s audited accounts show what under-used capacity costs, and the downstream brand-and-retail link, where the industry’s own gross-margin figures show where the profit sits. The rest is process: who holds title, who sets the margin, and where a domestic operator could insert a step that keeps value in the country.

A Ghanaian port warehouse worker moving a pallet of jute sacks of cocoa beans toward a cargo ship docked outside, other workers and stacked sacks nearby, documentary photograph

Structure, not price

Domestic margins are fixed by committee; the held evidence starts at the port.

2 of 9

links that carry real, held financial evidence: grinding and downstream

Table 7: The held figures this pillar is built on, and where each comes from
Measure	Figure	What it is and where it comes from
Value retained in Ghana	About 18 per cent of EU consumer-chocolate value	The share Ghana keeps of the value of the chocolate its beans end up in on European shelves; Naydenov et al. (Ecorys/RVO), 2022, p.5. The value-ladder anchor.
Farm-gate share of a bar	About 5 to 6 per cent of the retail price	The farmer's slice of a finished chocolate bar, pre-crisis; Cocoa Barometer 2025. The second value-ladder anchor.
Brand and trader margins	Premium brands above 65%; mainstream 35 to 50%; traders 8 to 20%	Gross profit margins along the downstream chain; Cocoa Barometer 2025. Where the profit concentrates, all of it outside Ghana.
Installed grinding capacity	About 514,000 t/yr, over 60 per cent of a harvest	Ghana-based processors' installed grinding capacity; Naydenov et al. 2022.
Grinding capacity used	About 50 per cent; ~259,000 t processed 2019	Actual grinding against installed capacity; Naydenov et al. 2022. The idle-plant gap.
State grinder result	Loss after tax US$12.06m, FY2022	Cocoa Processing Company's audited loss, blamed on bean unavailability; Cocoa Processing Company 2022.
Raw against processed export	About 70 per cent raw, 30 per cent processed	The split of the crop at export; Aning 2023. The value-addition gap stated as a flow.

Sources: Naydenov et al. (Ecorys/RVO), Cocoa Processing Study (2022); Fountain and Huetz-Adams, Cocoa Barometer 2025; Cocoa Processing Company audited accounts (2022); Aning, ICCO Ghana Country Report (2023). Each figure is cited again in the body prose where it is used.

A bar showing the share of a chocolate bar's value that stays in Ghana against the share captured beyond it, and a second chart of downstream gross margins for premium brands, mainstream brands and traders — Figure 10 Where the value of a bar of chocolate goes, and how little of it stays in Ghana.

What this shows

Built on the Cocoa Barometer’s farm-gate share (about 5 to 6 per cent of a finished bar) and downstream gross margins (premium brands above 65 per cent, mainstream brands 35 to 50 per cent, traders 8 to 20 per cent), and on the RVO/Ecorys finding that Ghana retains about 18 per cent of the value of consumer chocolate sold in the EU. The per-kilogram split some sources print is not reproduced; it is not in a source held for this playbook.

The master chain: nine links, and where the openings cluster

Each row below is opened out into its own section afterwards. Read the fourth column first. The breaks are where the openings are, and in this chain they cluster at one end. The structure of the nine links is the authors’ mapping against the sources named beneath the table; the marketing-structure mechanics are from Aning (2023) and Mirza and Pringle (2024), the grinding evidence from Naydenov et al. (2022) and Cocoa Processing Company (2022), and the downstream margins from Fountain and Huetz-Adams (2025).

Table 8: The nine links, who runs each, the margin signal, where value leaks, and the opening inside the leak
Link	Who operates it	Margin or price signal	Where value leaks or breaks	The opening inside the break
1. Input supply and nursery	COCOBOD's subsidised Hi-Tech scheme, a thin scatter of agrovet dealers, and CRIG-sourced hybrid seedlings	Inputs are subsidised, not priced by a market; the farmer rarely pays the true cost	Subsidised inputs are diverted and resold, and no dealer builds a commercial supply because the state gives inputs away	Sell the input bundled with a record: certified material and a documented application history a buyer will pay for
2. Farm production	More than 760,000 smallholder farmers on farms of a few hectares	Administered producer price, set once a season by the PPRC, targeting at least 70 per cent of the net FOB	The farmer is a price-taker by law, not by weakness; the price is fixed before the season opens	Lift the yield, not the price; the volume on the same land is yours, the price is not
3. LBC purchasing and clerks	About 30 licensed buying companies buying only through commissioned purchasing clerks; state PBC the largest	A fixed buyer's margin per tonne set by the PPRC; LBCs compete on volume, not price	LBCs cannot differentiate on price, so viability turns on seed finance and volume; the small ones fail when finance is late	Aggregate volume and carry your own finance; the buying layer rewards reliability, not a better offer
4. Grading and sealing (QCC)	Quality Control Company, a COCOBOD subsidiary, grading into Grade I, Grade II and substandard	No margin; a quality gate that protects the origin premium	The premium the grading protects is captured downstream, not by the grader or the farmer	Build traceability on top of the grading record that already exists; the certificate is the asset
5. Take-over (CMC)	Cocoa Marketing Company, a wholly-owned COCOBOD subsidiary, taking title at Tema, Takoradi and Kaase	Net FOB price; CMC is the single seller of Ghana's cocoa	Title, and with it all price exposure and export margin, passes to a state monopoly	None for a private entrant here; it is a state gate, worked around rather than entered
6. Haulage and warehousing	COCOBOD-approved private hauliers on PPRC-regulated rates	Regulated haulage rate per district (rate figure not published)	An unmeasured, regulated cost with no public rate card and no named operator roster	Efficiency and record-keeping, not rate negotiation; the rate is set, the reliability is not
7. Grinding and primary processing	Three multinationals (Cargill, Barry Callebaut, Olam) and indigenous grinders (Niche, CPC, Cocoa Touton)	Processing margin on liquor, butter, powder, cake	Installed capacity sits about half idle; the state grinder runs at a loss for want of beans	The capacity already exists; the constraint is bean supply and working capital, not plant
8. Export	CMC to external buyers; grinders export semi-finished products	Net FOB for beans, a higher unit value for processed products	About 70 per cent of the crop leaves raw, so the value added by grinding leaves with someone else	Move a tonne from the raw column to the processed column; the mandate now requires it
9. Downstream manufacture and retail	Brands and retailers, all outside Ghana	Retail price and brand gross margin	Brands take the largest margin in the chain, protected by intellectual property, not by cocoa	Not a link Ghana can enter at scale; the realistic prize is primary processing, link 7

Sources: Aning (2023) for the marketing structure, take-over and district figures; Mirza and Pringle (2024) for the LBC and purchasing-clerk mechanics; Naydenov et al. (2022) and Cocoa Processing Company (2022) for the grinding link; Fountain and Huetz-Adams (2025) for the downstream margins. The nine-link chain structure is the authors' mapping against those sources.

What this shows

Read this chain by where title moves and where value is added, and note they are not the same place. Title moves through a fully administered domestic chain, farmer to clerk to LBC to QCC to CMC, at prices fixed before the season starts, so nobody in Ghana is fighting over margin in links 1 to 6. Value is added at links 7 to 9, and Ghana only reaches the first of those three. The whole opportunity in this pillar is one sentence long: the country has built the grinding step and is not using it.

Link 1. Input supply and the nursery: subsidy instead of a market

The chain starts with a link the state has taken over and, in taking it over, has kept a commercial supply industry from forming. COCOBOD distributes fertiliser and agro-inputs to cocoa farmers under a heavily subsidised scheme, and in 2026 reinstated a fully government-funded free-input programme as part of its reform package.³The intent is access and affordability, and the effect on the value chain is that no private dealer builds a business selling to cocoa farmers, because the farmer’s alternative is a free or near-free state input.⁴

The weakness in that arrangement is leakage. COCOBOD itself has issued a public notice warning that subsidised fertiliser and agro-inputs were being illegally diverted and resold by unscrupulous persons, including farmers colluding in the practice, and has called on security agencies to intercept undocumented shipments.⁵ A subsidised input with no record attached to it is an input that can be diverted, and the same absence of a record is what stops the input from ever becoming a selling point. The commercial opening here is not to compete with a free state input on price, which is impossible, but to sell the one thing the state scheme does not attach: a documented history. Certified planting material from CRIG-sourced hybrids, sold with a written record of what was applied and when, is a product a traceability-conscious buyer will eventually pay for, and it is the same record the rest of this chain is missing.⁶

A Ghanaian agricultural extension worker distributing cocoa hybrid seedlings in black nursery bags to a smallholder farmer at a COCOBOD input distribution point, sacks of fertiliser stacked nearby, documentary photograph

Free, but with no record attached

Free inputs with no record attached; the record, not the seedling, is what the chain is missing.

Link 2. Farm production: the price-taker who is one by law

The producer sells at a price set once a season by the Producer Price Review Committee, a government body, with the stated objective of paying farmers at least 70 per cent of the net free-on-board price.⁷This is the defining feature of Ghana’s cocoa chain and the reason a value-chain analysis of it reads differently from any other commodity in this series. In the small-ruminant chain, or the maize chain, the farm-gate price is negotiated animal by animal or bag by bag, and the producer’s weakness is a weakness of scale and information. In cocoa, the farm-gate price is administered. The more than 760,000 smallholder farmers⁸who grow Ghana’s cocoa are not price-takers because they lack bargaining power; they are price-takers because there is nothing to bargain over.

The consequence for anyone reading this chain for an opening is that the entire price story at the farm is the wrong place to look for one. The price is the same whether a farmer sells a poor crop or an excellent one, so the only lever a farmer or an aggregator controls at this link is volume: more beans off the same land at the same administered price. That is a production question, owned by Pillars 1 and 4, and it is the reason this pillar’s openings sit downstream of the farm rather than at it. The one qualification is quality grade, treated at link 4, which the farmer influences through fermentation and drying but is not directly paid a market premium for.

Link 3. LBC purchasing: a buying layer that cannot compete on price

About 30 licensed buying companies operate in Ghana, and none of them may buy a bean directly from a farmer. Each buys through locally established purchasing clerks, who work on commission, and each is reimbursed by COCOBOD for what it pays the farmer plus a fixed buyer’s margin per tonne, a margin set by the PPRC at the beginning of every season.⁹Because the margin is fixed and identical across companies, LBCs cannot compete on the price they offer the farmer or the margin they earn. They compete only on volume sourced and on the quality of the service, the training and the field officers, that keeps a purchasing clerk and the clerk’s farmers loyal.¹⁰

That structure decides which LBCs survive, and the answer is the ones with finance and volume. The state’s Cocoa Marketing Company supplies the seed finance that LBCs use to pay farmers, and when that finance arrives late, as it did in 2022, small and independent LBCs cannot open their sheds while the multinational-backed ones self-finance and keep buying. The squeeze is visible in a single company’s record: the case LBC studied by IISD saw its purchasing fall from 53,000 tonnes in 2018/19 to 44,000, then 32,000, then 22,000 tonnes by 2021/22, a collapse driven by finance and by competition from larger rivals rather than by any failure of the farmers it worked with.¹¹ For a domestic entrant the lesson is that this link rewards two things the fixed margin cannot: reliable volume and the working capital to buy without waiting for COCOBOD. An aggregation model that carries its own finance is competing on the one dimension the administered margin leaves open.¹²

A Ghanaian licensed buying company clerk writing in a ledger at a rural purchasing shed, a farmer waiting beside a hanging scale and stacked jute sacks, documentary photograph

Financed, not merely licensed

53,000 to 22,000 tonnes in four seasons: one LBC's record of what late seed finance does to volume.

What this shows

Do not plan to out-price the buying layer, because the price is fixed and there is nothing to out-price. Plan to out-last it. The LBCs that fail are the ones that run out of cash when the state’s seed finance is late, and the ones that survive are financed and can fill volume. An entrant’s edge at this link is a balance sheet and a collection network, not a better offer to the farmer.

Link 4. Grading and sealing: the gate that protects a premium Ghana does not keep

A Ghanaian quality control grading officer inspecting and sealing a jute sack of dried cocoa beans at an upcountry grading depot, other graded and stamped sacks stacked nearby, documentary photograph

The gate works; the premium leaves anyway

The grading works and the premium is real; it is captured downstream, not at this gate.

Every bag of Ghanaian cocoa passes through the Quality Control Company, a COCOBOD subsidiary, which grades it into Grade I, Grade II or substandard and seals it before it moves. QCC inspects at the upcountry grading centres, again at the ports, and a final time before shipment, with staff across 73 districts, and its stated purpose is to protect Ghana’s reputation for well-fermented, defect-free beans that commands a premium on the world market.¹³The grading works, and Ghana’s origin premium is real. The point for this pillar is where that premium is captured, and the answer is not at the gate that protects it. The grader earns no margin, the farmer earns the administered price whatever the grade, and the premium for clean, traceable, single-origin Ghanaian cocoa is realised by the brands that put ‘Ghana’ on a wrapper.¹⁴

The opening here is not the grading, which the state already does well, but the record the grading generates. A bean that has been graded, sealed and logged at a named centre already carries most of the information a traceability system needs. The gap, treated in the digital section below, is that the record stops at the sack and does not follow the bean to the farm it came from or forward to the buyer who wants to prove where it grew.

Link 5. Take-over: the state monopoly at the centre of the chain

Title to Ghana’s cocoa passes to the state at a single point. The Cocoa Marketing Company, a wholly-owned COCOBOD subsidiary, takes over the graded and sealed beans from the LBCs at take-over centres at Tema, Takoradi and the inland terminal at Kaase near Kumasi, and from that moment the cocoa is the state’s to warehouse, market and export.¹⁵CMC is the sole seller of Ghana’s cocoa to the world, and it sets the export contracts, including the forward sales that fix a large part of each crop’s price before the beans are even bought.¹⁶

This link is not an opening for a private operator, and the honest course is to say so plainly rather than imply a gap that is not there. The take-over is a state gate by design. What matters for an entrant is that everything upstream of it, links 1 to 4, is a controlled feeder system for a monopoly, and everything downstream of it, links 7 to 9, is where private value can be added. The take-over is the hinge between the administered half of the chain and the competitive half.

Link 6. Haulage and warehousing: a regulated cost with no visible rate

Beans move from district depots to the take-over centres on COCOBOD-approved private hauliers, at rates the PPRC regulates per district according to distance and road condition.¹⁷ The haulage layer is genuinely private, but it is priced like the rest of the administered chain, by committee rather than by the market. Named operators do surface, Tarzan Enterprises and Global Haulage among them,¹⁸ but no source held for this playbook records an actual haulage rate, in cedis per tonne or per kilometre; the rate exists and is regulated, but its level is not published. That absence is worth stating because it marks the limit of what can be claimed about this link: it is a real cost, borne on every tonne, that this pillar cannot price. The opening, such as it is, is operational rather than commercial, in reliability and record-keeping, because the rate is fixed and there is no margin to negotiate.

Link 7. Grinding: the step Ghana built and does not use

This is the link where the value-addition story is decided, and it is the one link in the domestic chain with hard financial evidence behind it. Ghana-based processors hold enough installed capacity to grind about 514,000 metric tonnes of cocoa a year, equivalent to over 60 per cent of an average harvest, and only about half of that capacity is used.¹⁹That figure sits close to the 504,780 tonnes of installed capacity USDA reports and Pillar 2 uses, so the two pillars are describing the same plant from two sources rather than two different totals. In 2019 the country processed roughly 259,000 tonnes into liquor, powder, butter and paste. The plants exist. Ghana already has enough installed grinding capacity to meet the government’s target of processing half the crop; what it does not have is the beans, the finance and the utilisation to run those plants at their rated output.²⁰

A wide interior view of a Ghanaian cocoa processing factory with large industrial grinding and pressing machines, most of the plant quiet and idle, dim overhead lighting, no workers visible, documentary photograph

Idle plant, not absent plant

The plant is built and standing; what it lacks is beans to run through it.

514,000 t

installed grinding capacity, over 60% of an average harvest

~259,000 t

actually processed in 2019, about half of installed capacity

A bar chart of installed grinding capacity against cocoa actually processed, showing installed capacity at 504,780 tonnes and actual processing at 259,000 tonnes, about half of installed capacity — Figure 11 Installed grinding capacity against capacity actually used.

What this shows

Ghana holds about 514,000 tonnes of installed capacity, over 60 per cent of an average harvest, and runs it at roughly 50 per cent utilisation, about 259,000 tonnes processed in 2019. The gap between the two bars is idle plant, not absent plant; the state grinder’s own accounts show what that idle half costs.

The clearest evidence of what under-used grinding capacity costs comes from the state’s own processor. The Cocoa Processing Company, which makes the GoldenTree brand, is 57.73 per cent owned by COCOBOD, with the Ministry of Finance and SSNIT holding special preference shares alongside it, so it is a state-owned grinder in the most direct sense. Its audited accounts for the year to September 2022 show a loss after tax of US$12.06m, following a US$15.09m loss the year before, and current liabilities that exceeded current assets by US$69.34m. The auditors recorded a material uncertainty over the company’s ability to continue as a going concern, the company had defaulted on principal and interest to its bank syndicate, and COCOBOD had converted US$32m of loans and US$55m of trade payables into a claim on shares to keep it standing.²¹²²

The reason the company gives for the loss is the whole thesis of this pillar in one line from its own directors: the loss-making position is significantly driven by the unavailability of cocoa beans to process. The grinder is not losing money because Ghanaians will not buy chocolate or because the plant is obsolete. It is losing money because it cannot get beans to run through a plant it has, and its cocoa-processing segment turned in a gross loss of US$5.88m while its smaller confectionery segment made a modest profit. The company earned no tolling revenue at all in 2022, meaning it was not even paid to grind beans on behalf of others. Its response to the crisis was to borrow further: a bank syndicate led by Absa funded an expansion of capacity from 25,000 to 65,000 tonnes, and a US$86.7m Afreximbank facility was sought to retool plant and equipment, in a plant it could not fill at its existing size.²³²⁴

A quiet Ghanaian state-owned cocoa processing factory floor, tall grinding and pressing machinery standing idle, a lone manager reviewing a clipboard, documentary photograph

The plant exists, the beans do not

The state grinder lost money not for want of a plant, but for want of beans to run through it.

US$12.06m

Cocoa Processing Company's audited loss after tax, FY2022

US$69.34m

the shortfall of current assets against current liabilities

What this shows

The binding constraint on Ghanaian cocoa value-addition is not a missing factory. It is bean supply and working capital reaching the factory that exists. A state grinder that cannot get beans and cannot pay its banks, then borrows to add capacity it already cannot fill, is a warning about sequencing: the money that would build new plant is better spent securing a bean supply and financing throughput at the plant already standing. For a private operator the same logic points to tolling and contract grinding against secured beans, not to a new build.

Link 8. Export: the raw column and the processed column

About 70 per cent of Ghana’s cocoa leaves the country as raw beans, and about 30 per cent is processed domestically into semi-finished products before it leaves.²⁵Those two numbers are the value-addition gap stated as a flow. Every tonne in the raw column earns Ghana the net FOB bean price and nothing more; every tonne moved into the processed column earns the higher unit value of liquor, butter or powder, and keeps the grinding margin, the jobs and the foreign exchange in the country. The government’s response is a mandate to lift domestic processing from about a third of the crop today to a minimum of half from the 2026/27 season.²⁶The capacity to meet that mandate already exists, as link 7 showed; whether the beans and the finance will reach it is the open question, and it is the same question that sank the state grinder’s accounts.

Link 9. Downstream manufacture and retail: where the margin actually is

A macro photograph of two dark chocolate bars, one broken in half showing the glossy cross-section, resting on a plain wooden board, documentary photograph

Where the margin actually is

Everything that makes this valuable happened after Ghana was already paid.

US$10bn/yr

the cocoa sector's living-income gap, per Cocoa Barometer

US$10bn/yr

returned to chocolate-company shareholders in the same period

The far end of the chain is where the profit concentrates, and none of it is in Ghana. The industry’s own figures, compiled in the Cocoa Barometer, show premium chocolate brands such as Lindt earning gross margins above 65 per cent, mainstream brands such as Mondelez, Hershey and Nestle earning 35 to 50 per cent, and the cocoa traders that sit between farm and brand earning only 8 to 20 per cent. The brands earn the most because they capture profit through their intellectual property with minimal investment, and their margins have stayed stable for a quarter century regardless of whether cocoa prices rose or fell, because they hedge the cocoa price and pass input costs through to the shelf. The scale of the imbalance is captured in one comparison from the same report: the cocoa sector’s living-income gap, the shortfall between what farmers earn and a decent living, was estimated at around US$10bn a year, while chocolate companies returned more than US$10bn a year to their own shareholders through dividends and buybacks.²⁷²⁸

This link is not one Ghana can realistically enter at scale, and a serious value-chain analysis should not pretend otherwise. Building a global chocolate brand from an origin country is a rare achievement, and the honest prize for Ghana is one step back, at link 7, capturing more of the grinding and primary-processing margin rather than chasing the branded-retail margin that sits behind decades of marketing spend. The value ladder is steep, and the reachable rung is the grinding one.

The missing middle, named and diagnosed

Put the nine links back together and the shape is clear. Ghana runs a long, well-organised, fully administered chain from the farm to the port, and then the chain stops. The country grows the beans, grades them, seals them, and hands them to a state monopoly that sells most of them raw. The value that other countries add, the grinding, the blending, the branding, the retail, sits in links 7 to 9, and Ghana reaches only the first of those, and reaches it at about half capacity. The missing middle in cocoa is not aggregation, as it is in the livestock chains, because COCOBOD’s marketing system already aggregates every bean in the country. The missing middle is domestic value-addition: the grinding and primary processing that is built but idle, and the working capital and bean supply that would let it run.

Two forces keep that middle empty. The first is finance. The whole chain runs on borrowed money, from the farmer paying up to 100 per cent interest over six months to private lenders, to the LBCs waiting on late seed finance, to the state grinder defaulting on its syndicate.²⁹ Ghanaian banks themselves name inventory finance for seed-crop purchase and asset finance for buyers and haulage as the sector’s core credit needs.³⁰ International finance is now moving toward the buying layer, with the IFC and Societe Generale Ghana structuring a facility to mobilise up to US$80m for LBCs to buy directly from farmers, but the processing layer, where the value is, remains starved.³¹The second force is the administered price system itself. As van Huellen and colleagues argue, COCOBOD’s monopoly on cocoa sales is constrained by its dependence on international financing and derivative-market price references, which is why trading-company margins persist even under a state marketing board.³² The country that most tightly controls its cocoa chain still cannot control where in that chain the value ends up.

The one domestic model that reaches past the port is farmer ownership. Kuapa Kokoo, a cooperative of around 100,000 members, owns its own licensed buying company, co-owns the Divine Chocolate brand, and in 2024 committed GH¢5m to begin producing semi-finished and finished cocoa products in Ghana.³³ It is the clearest existing example of Ghanaian value-chain integration, and it is the exception that proves the rule: reaching the grinding link and beyond took a farmer institution three decades to build, and it remains small against the volume the multinationals grind at Tema.³⁴

Who holds power over this chain, and who wants the middle built

A stakeholder map earns its place only if its axes change a decision. In this chain the useful axis is not interest in the sector, since everyone in cocoa is interested in cocoa. The useful axis is interest in shifting value downstream into domestic processing, because that is the change this pillar identifies, and the actors sort very differently on it. COCOBOD has near-total power over the chain and a genuine, stated interest in domestic processing, but its own state grinder shows how hard that interest is to execute. The multinational grinders have the plant and the working capital to process but no reason to move value-capture from their own global operations to Ghana. And the farmer, who would gain most from a deeper domestic chain, has the least power to build one.

A power-interest grid placing COCOBOD, the Cocoa Marketing Company, multinational grinders, licensed buying companies, the Cocoa Processing Company, international finance, farmer cooperatives, brands and retailers, and smallholder farmers, by power over the chain and by interest in keeping value in Ghana — Figure 12 The chain’s actors placed by power over the chain and by appetite for keeping value in Ghana.

What this shows

The useful axis is not interest in cocoa, since everyone in cocoa is interested in cocoa, but interest in shifting value downstream into domestic processing. COCOBOD has the power and the stated interest yet its own grinder cannot execute; the multinational grinders have the plant and no reason to move value-capture to Ghana; the farmer would gain most and has the least power to build it.

Table 9: The chain's actors, sorted by power and by appetite for keeping value in Ghana
Actor	Role in the chain	Power	Interest	What it means for an entrant
COCOBOD	Sets the price, owns the marketing monopoly, owns the state grinder	High	High	The gatekeeper for everything. Its mandate and finance make the market, yet its own grinder's accounts show it cannot execute processing alone. Align with the mandate; do not wait for COCOBOD to operate
Cocoa Marketing Company	Sole seller and exporter of Ghana's cocoa	High	Medium	The state gate you cannot enter. Your beans reach the world through it; work with its take-over system, not around it
Multinational grinders	The working grinding capacity at Tema	High	Low	They hold the plant and the finance and capture value in their own global chains. Potential tolling partners, not allies in keeping margin in Ghana
Licensed buying companies	Buy from farmers on a fixed margin, including the state PBC	Medium	Medium	Financed, volume-driven, and squeezed when seed finance is late. The layer to join or aggregate into, if you carry your own capital
Cocoa Processing Company	The state's domestic value-addition attempt	Low	High	Wants processing to work and cannot fund it. A cautionary case and a possible contract-grinding counterpart, not proof the model pays
International finance	Working capital for the chain (IFC, GIRSAL, banks)	High	High	Moving money toward the buying layer today. Records and contracted volume are the entry ticket; the processing layer is still under-served
Farmer cooperatives	Farmer-owned buying and, rarely, processing (e.g. Kuapa Kokoo)	Low	High	The one model that reaches past the port. Slow to build, but the template for domestic integration and a ready partner in traceability
Brands and retailers	Downstream manufacture and retail	High	Low	Capture the largest margin, outside Ghana, protected by IP. Not a link to enter; the reason the value ladder is steep
Smallholder farmers	Grow the crop, capture the least (more than 760,000 farmers)	Low	High	The supply base and the constituency for a deeper chain. Organising them is the highest-impact and slowest move available

Sources: Aning (2023) and Mirza and Pringle (2024) for the marketing-structure roles; Cocoa Processing Company (2022) for the state grinder; Fountain and Huetz-Adams (2025) for the downstream actors; IFC (2025) and International Trade Centre (2024) for finance and cooperatives. Power and interest positions are the authors' assessment on that evidence.

Market signal

Watch the 50 per cent processing mandate and the money behind it, not the announcement. A mandate to grind half the crop from 2026/27 is only as real as the bean supply and working capital that reach the grinders, and the state’s own processor has already shown that a directive without finance produces a loss, not a product. If the reform package pairs the mandate with secured beans and processing finance, the grinding link opens; if it does not, the capacity stays idle and the mandate slips.

Digital and AI in the chain: traceability, and why the record comes first

A Ghanaian cocoa farmer and a field agent standing at the edge of a cocoa farm, the agent attaching a small tag to a tree trunk while holding a tablet computer, cocoa pods visible on nearby trees, documentary photograph

A record, not a model

A tagged parcel and a mapped farm, not an algorithm, is the precondition for all of this.

Three digital layers would change this chain, and they sit at very different distances from being real in Ghana. The failure mode to avoid is describing a foreign platform as though it ran here, or recommending a farmer a system only a company can afford. Traceability is the nearest and the most consequential, because the chain’s central weakness is a record that stops at the sack. About 60 per cent of the world’s cocoa was still untraced in 2024/25, and the beans that pass through Ghana’s grading system acquire a grade and a seal but not a link back to the farm they grew on.³⁵The Ghana Cocoa Traceability System, piloted with COCOBOD, is the domestic effort to close that gap, and the satellite farm-mapping already in this playbook’s evidence base through Kalischek’s forest-boundary work is the raw material a traceability system needs.³⁶ The precondition here is a record, not a model. Ghana does not first need an algorithm; it needs every graded parcel tied to a mapped farm and a purchasing-clerk transaction, which is data the QCC grading step and the LBC purchasing step already generate and simply do not join up. Once that record exists, the EUDR compliance the sector faces from December 2026, and the origin premium the brands capture, both become things a Ghanaian operator can prove and charge for.

The second layer is a farmer and cocoa-management database, the work COCOBOD has undertaken to know who its farmers are and where their farms sit. This is not a value-capture tool on its own, but it is the spine every other digital layer hangs from, because a payment, a subsidy or a traceability claim is only as good as the farmer record behind it, and the input-diversion problem at link 1 is at root a record problem. The third layer, downstream analytics and demand matching of the kind the brands run, is real but sits with the actors Ghana cannot reach, and it is honest to leave it there. The digital opening for a Ghanaian entrant is the first layer, built cheaply: a tagged parcel, a mapped farm, a logged transaction, and the certificate that comes out the other end.

The value-capture openings, ranked

The order below runs from the opening with the strongest held evidence and the lowest capital requirement to the largest prize and the longest odds. Each row states what it requires, not what it would achieve, and each rests on a source cited in this pillar’s prose.

Table 10: Five openings in the cocoa chain, ordered by evidence and by what each costs to start
Rank and opening	What the evidence says	What it requires	Who it suits
1. Traceability and origin-record services	About 60 per cent of world cocoa is untraced, EUDR compliance lands from Dec 2026, and Ghana's grading and purchasing steps already generate the raw records	A way to join the QCC grade record, a mapped farm and an LBC transaction into one certificate; software and field data, not plant	A digital or agribusiness operator with field access, working with COCOBOD's traceability pilot
2. Aggregation with carried finance	LBCs compete only on volume and fail when the state's seed finance is late; the fixed margin rewards reliability	An LBC licence or a partnership with one, independent working capital, and a purchasing-clerk network	A financed trader or cooperative, not a thinly capitalised newcomer
3. Contract grinding and tolling	Installed grinding capacity runs at about 50 per cent, and the state grinder earned no tolling revenue at all in 2022	A secured bean supply, working capital for throughput, and a tolling contract with an existing plant	An operator who can finance beans, rather than one who wants to build a factory
4. Farmer-owned processing integration	Kuapa Kokoo reached its own LBC and a processing commitment over three decades; it is the only domestic model that passes the port	A cooperative base, patient capital, and a decade-plus horizon	A farmer institution or an impact investor, not a quick-return entrant
5. A new primary-processing build	Capacity already exceeds the 50 per cent mandate, so a new plant competes for the same scarce beans that idle the existing ones	Everything in row 3, plus the capital for plant, and a bean supply the existing grinders do not already claim	A large industrial investor, and only once bean supply and finance are proven

Sources: Fountain and Huetz-Adams (2025) for the traceability and value figures; Mirza and Pringle (2024) for the LBC finance mechanics; Naydenov et al. (2022) and Cocoa Processing Company (2022) for the grinding capacity and utilisation; International Trade Centre (2024) for the cooperative model. The ranking is the authors' judgement on that evidence.

The bottom line

The largest prize, a new processing build, sits at the bottom of that list for a reason the state grinder makes plain: the constraint is not plant, it is beans and finance, and adding plant against a scarce bean supply reproduces exactly the idle-capacity problem that already exists. The strongest opening is the least capital-intensive, the record itself, because it is the one thing this chain is missing at every link and the one thing every downstream premium is now going to demand.

The risks that sit inside the chain

A Ghanaian cocoa processing factory worker feeding beans into a grinding machine, other machinery visible around him, documentary photograph

Idle grinding capacity and the under-fed plant

HIGH

What it is

Ghana has built enough grinding capacity to process over 60 per cent of its harvest and runs it at about half, so the country's one domestic value-addition step operates far below what it could earn.

Evidence

Installed capacity is around 514,000 tonnes against roughly 259,000 tonnes actually processed, and the state-owned grinder ran at a loss two years running, blaming the unavailability of beans to process, defaulting on its banks, and earning no tolling revenue at all in one of those years.

Who it hits

The national value-addition ambition, the workers in idle plants, and any investor who assumes the constraint is a missing factory rather than a missing bean supply.

How to manage it, and the opening

Treat bean supply and working capital as the scarce inputs, not plant. Contract grind against secured beans before building anything, and read the 50 per cent mandate as a demand signal that will only convert to product if finance follows it.

Late seed finance and the collapse of the independent buying layer

HIGH

What it is

The licensed buying companies depend on the state releasing seed finance on time, and when it comes late the independent LBCs cannot buy while the multinational-backed ones self-finance and take the volume.

Evidence

One licensed buyer's purchasing fell from 53,000 tonnes to 22,000 over four seasons as finance tightened and larger rivals competed, and the state's own largest buying company has been reported near insolvency with negative equity.

Who it hits

The smaller and farmer-owned buying companies first, then the farmers whose clerks cannot pay them at harvest.

How to manage it, and the opening

An entrant in the buying layer must carry independent working capital and not rely on the state release; the companies that survive this link are financed, not merely licensed.

Value locked downstream and outside Ghana

HIGH

What it is

The profit in cocoa concentrates in grinding, branding and retail, almost all of it outside the country, so Ghana can grow more and better cocoa without earning much more from it.

Evidence

Ghana retains about 18 per cent of the value of the chocolate its beans end up in on European shelves, the farm-gate price is only about 5 to 6 per cent of a finished bar, and premium brands earn gross margins above 65 per cent while the traders between farm and brand earn a fraction of that.

Who it hits

The whole national economy, and any strategy that treats higher yield or a higher administered price as the route to higher value.

How to manage it, and the opening

The reachable rung is primary processing inside Ghana, not the branded-retail margin behind decades of marketing; move tonnes from the raw column to the processed column rather than chasing the shelf.

A Ghanaian agro-input dealer writing in a ledger beside stacked fertiliser sacks and cocoa seedlings in nursery bags, documentary photograph

Input diversion and the missing record

MEDIUM

What it is

Subsidised state inputs are diverted and resold, and the same absence of a record that lets them be diverted is what stops the chain from proving origin or capturing a traceability premium.

Evidence

COCOBOD has publicly warned that subsidised fertiliser and agro-inputs were being illegally diverted and resold, including by farmers colluding in the practice, and about 60 per cent of the world's cocoa remained untraced in the most recent season.

Who it hits

The integrity of the subsidy, the farmers who miss the inputs, and every downstream buyer who now needs to prove where a bean grew.

How to manage it, and the opening

Attach a record to the input and the parcel; the farm database and traceability pilot already under way are the fix, and the certificate they produce is itself the sellable product.

Key takeaways

01

Ghana's cocoa chain is administered from the farm to the port and competitive only after it, and the country reaches only the first competitive link. The producer price is fixed by committee, so no value is fought over in the domestic chain.

02

The value is added downstream in grinding, branding and retail, where Ghana keeps about 18 per cent of what its cocoa is finally worth and the farm-gate is only about 5 to 6 per cent of a bar. Premium brands earn gross margins above 65 per cent, traders 8 to 20 per cent.

03

The one domestic value-addition step, grinding, is built to over 60 per cent of the harvest and run at about half. The state's own grinder, Cocoa Processing Company, ran a loss after tax of US$12.06m in 2022 because it could not get beans, not because it lacks a factory.

04

The binding constraint on keeping cocoa value in Ghana is bean supply, working capital and a record that follows the bean, not more plant and not a higher administered price. The reachable openings, in order, are the traceability record, financed aggregation and contract grinding; a new build sits last.

Written for each reader

Practitioner intelligence

Hover any card to pause and lift it.

For students

Do not analyse this chain by looking for where the farmer is underpaid at the gate, because the gate price is set by a government committee and is the same for everyone. Analyse it by asking where value is added and how much of that Ghana keeps, and the answer, about 18 per cent of the finished chocolate's value, tells you the whole story. The mistake specific to cocoa is treating it like an open-market crop where a better price can be negotiated; in Ghana the price is administered and the analysis belongs downstream of the port, not at the farm gate.

For entrepreneurs

The reachable opening is a record, not a refinery. Every downstream buyer is moving toward proof of origin, EUDR compliance lands from the end of 2026, and Ghana's grading and purchasing steps already generate the raw data and simply do not join it up. Build the certificate before you build anything physical. If you do move into processing, toll against a secured bean supply rather than building plant, because the constraint that idles the existing grinders, and bankrupted the state one, is beans and cash, not factories.

For investors

Value in Ghanaian cocoa is not stolen from the farmer, it is added abroad, so the investable asset is a step that moves value-capture back into Ghana rather than a larger share of the administered farm-gate trade. Ask any processing proposal where its beans come from and how the throughput is financed, because the state grinder had the plant and failed on exactly those two questions; ask any buying-layer proposal how it is capitalised independently of the state's seed-finance release; and treat a new-build grinding pitch with suspicion, since installed capacity already exceeds the national mandate and a new plant competes for the same scarce beans.

For ecosystem actors

The highest-impact public move is not another grinding plant but the bean supply and processing finance that would fill the plants already standing, paired with the traceability and farmer-record systems that let Ghana prove and charge for its origin. Fund the throughput and the record, not the concrete, and read the state grinder's accounts as the evidence that a processing mandate without finance produces a loss, not a product.

For students

Do not analyse this chain by looking for where the farmer is underpaid at the gate, because the gate price is set by a government committee and is the same for everyone. Analyse it by asking where value is added and how much of that Ghana keeps, and the answer, about 18 per cent of the finished chocolate's value, tells you the whole story. The mistake specific to cocoa is treating it like an open-market crop where a better price can be negotiated; in Ghana the price is administered and the analysis belongs downstream of the port, not at the farm gate.

For entrepreneurs

The reachable opening is a record, not a refinery. Every downstream buyer is moving toward proof of origin, EUDR compliance lands from the end of 2026, and Ghana's grading and purchasing steps already generate the raw data and simply do not join it up. Build the certificate before you build anything physical. If you do move into processing, toll against a secured bean supply rather than building plant, because the constraint that idles the existing grinders, and bankrupted the state one, is beans and cash, not factories.

For investors

Value in Ghanaian cocoa is not stolen from the farmer, it is added abroad, so the investable asset is a step that moves value-capture back into Ghana rather than a larger share of the administered farm-gate trade. Ask any processing proposal where its beans come from and how the throughput is financed, because the state grinder had the plant and failed on exactly those two questions; ask any buying-layer proposal how it is capitalised independently of the state's seed-finance release; and treat a new-build grinding pitch with suspicion, since installed capacity already exceeds the national mandate and a new plant competes for the same scarce beans.

For ecosystem actors

The highest-impact public move is not another grinding plant but the bean supply and processing finance that would fill the plants already standing, paired with the traceability and farmer-record systems that let Ghana prove and charge for its origin. Fund the throughput and the record, not the concrete, and read the state grinder's accounts as the evidence that a processing mandate without finance produces a loss, not a product.

Where this connects.The administered producer price is Pillar 2’s subject, its February 2026 reform is Pillar 8’s, and the yield gap that decides how many beans reach the chain in the first place is Pillar 4’s subject, which matters here because bean supply, not plant, is the binding constraint on the grinding link. The cost model, the sensitivity analysis and the markers that turn any processing or aggregation opening into a budget are built in Pillar 5, which is where the margins gestured at in this pillar become an enterprise. And the EUDR exposure that makes the traceability opening urgent, landing on the same southern regions where cocoa concentrates, is Pillar 3’s territory.

Footnotes

Naydenov, Delera, Boisseau, and Gaddari (Ecorys), Cocoa Processing Study: Final Report (Rotterdam: Ecorys, commissioned by the Netherlands Enterprise Agency (RVO), 2022).↑
Antonie C. Fountain and Friedel Huetz-Adams, Cocoa Barometer 2025 (VOICE Network, 2025).↑
Joshua Worlasi Amlanu, “Free Fertiliser Returns as COCOBOD Resets Farmer Support,” The Business & Financial Times Morning Edition, no. 4,846, July 14, 2026.↑
Amlanu, “Free Fertiliser Returns.”↑
Ghana Cocoa Board (COCOBOD), Diversion of COCOBOD Fertilizers and Agro-Inputs (Accra: Ghana Cocoa Board, COCOBOD News, 2022).↑
COCOBOD, Diversion of COCOBOD Fertilizers and Agro-Inputs.↑
Alexis Fritz Kwabena Aning, Feasibility Study on Africa Cocoa Exchange (AfCX), Appendix II: Value Chain and Sectoral Regulatory Analysis, Country Reports, Ghana Country Report (Abidjan: International Cocoa Organization, 2023).↑
Ghana Cocoa Board (COCOBOD), Cocoa Sector Turnaround Strategy Paper (Accra: Ghana Cocoa Board, 2023).↑
Hafiz Mirza and Duncan Pringle, A Licensed Buying Company in Ghana's Cocoa Industry (Winnipeg: International Institute for Sustainable Development, 2024).↑
Mirza and Pringle, A Licensed Buying Company in Ghana's Cocoa Industry.↑
Mirza and Pringle, A Licensed Buying Company in Ghana's Cocoa Industry.↑
Mirza and Pringle, A Licensed Buying Company in Ghana's Cocoa Industry.↑
Quality Control Company Ltd (QCC), Grading and Sealing (Accra: Ghana Cocoa Board, n.d.).↑
QCC, Grading and Sealing.↑
Aning, Feasibility Study on Africa Cocoa Exchange (AfCX): Ghana Country Report.↑
Aning, Feasibility Study on Africa Cocoa Exchange (AfCX): Ghana Country Report.↑
Aning, Feasibility Study on Africa Cocoa Exchange (AfCX): Ghana Country Report.↑
GCB Bank, Strategy and Research Department, Cocoa Sector Analysis 2023 (Accra: GCB Bank, 2023).↑
Naydenov et al., Cocoa Processing Study.↑
Naydenov et al., Cocoa Processing Study.↑
Cocoa Processing Company PLC, Annual Report and Financial Statements for the Year Ended 30 September 2022 (Tema: Cocoa Processing Company PLC, 2022).↑
Cocoa Processing Company, Annual Report and Financial Statements 2022.↑
Cocoa Processing Company, Annual Report and Financial Statements 2022.↑
Cocoa Processing Company, Annual Report and Financial Statements 2022.↑
Aning, Feasibility Study on Africa Cocoa Exchange (AfCX): Ghana Country Report.↑
Ghana Cocoa Board (COCOBOD), Cocoa Sector Reforms: Raising Domestic Processing (Accra: Ghana Cocoa Board, 2026).↑
Fountain and Huetz-Adams, Cocoa Barometer 2025.↑
Fountain and Huetz-Adams, Cocoa Barometer 2025.↑
Mirza and Pringle, A Licensed Buying Company in Ghana's Cocoa Industry.↑
GCB Bank, Cocoa Sector Analysis 2023.↑
International Finance Corporation (IFC), IFC and Societe Generale Ghana Partner to Support Ghana's Cocoa Sector (Washington, DC: International Finance Corporation, 2025).↑
Sophie van Huellen et al., “The Political Economy of Pricing and Ghana's Cocoa Marketing System,” Competition & Change published online ahead of print, June 24, 2026.↑
International Trade Centre (ITC), Ghana's Kuapa Kokoo Farmer Cooperative Unveils a Strategic Plan for Cocoa (Geneva: International Trade Centre, 2024).↑
ITC, Kuapa Kokoo Unveils a Strategic Plan.↑
Fountain and Huetz-Adams, Cocoa Barometer 2025.↑
Nikolai Kalischek et al., “Cocoa Plantations Are Associated with Deforestation in Côte d'Ivoire and Ghana,” Nature Food 4, no. 5 (2023): 384-93.↑