Ghana’s cocoa crisis: who finances the wait?

A guaranteed price does not guarantee a payment date. That distinction sits at the centre of Ghana’s cocoa problem. A government can legislate for a better share of export earnings and a public agency can repay bondholders without answering the question that matters to a grower: how long after delivering the beans will the money arrive?

Recent announcements suggest progress. Ghana’s presidency reported the signing of the new Ghana Cocoa Board Act on 26 August 2026. On 1 September, COCOBOD, the public body overseeing the industry, said it had paid about GHS2.306 billion to holders of bonds affected by the domestic debt restructuring. Neither event, however, provides a record of payments received by individual farmers. 1 2

The useful story begins well before the chocolate factory. Someone must finance the purchase of a crop while its sale proceeds are still in the future. Following that money reveals why an industry can have beans to sell, a positive expected margin and too little cash to buy the next delivery.

A lower purchase price did not automatically unlock credit

On 19 March, Reuters reported that licensed cocoa buyers were still short of funds to purchase beans. Some available money was going towards earlier obligations. This is evidence of a financing blockage in March, not proof that precisely the same conditions persisted into September. 3

Official price decisions show how large the adjustment had already been. The farmgate rate announced in August 2025 was GHS51,660 per tonne. It rose to GHS58,000 from 3 October, then fell to GHS41,392 on 12 February 2026. That rate was maintained for the light crop opening on 18 June. The October-to-February reduction was 28.6% in nominal local-currency terms. It was not a change in the world cocoa price or a measure of farmers’ real income. 4 5 6

Sources: Ministry of Finance, 12 February briefing, COCOBOD, 12 June. First bar: 4 August announcement; others: effective dates. June covers the 2025/26 light crop, not 2026/27. The October–February cut was 28.6%.

Data and notes

Announcement / effective date	GHS / t	Publication
August 2025 announcement	51,660	2025-10-02
3 October 2025	58,000	2025-10-02
12 February 2026	41,392	2026-02-12
18 June 2026	41,392	2026-06-12

Cutting the purchase price makes the next tonne cheaper to finance. It does not settle the bill for the previous one. Nor does it oblige a bank to renew a credit line. When incoming funds first have to clear old obligations, a lower price may produce less new purchasing activity than expected.

The IMF’s July assessment describes the combination that squeezed the 2025/26 season: lower international prices, elevated domestic procurement costs, unsold stocks and acute cash-flow pressure. That provides an institutional assessment beyond COCOBOD’s own account. It does not establish a single cause for every delayed payment. 7

A pricing rule determines how much the grower is owed. Available finance, sales collections and existing obligations determine whether that amount can be paid on time. Changing the first does not automatically repair the second.

Who finances the beans before the customer pays?

Ghana’s system involves growers, Licensed Buying Companies, known as LBCs, and COCOBOD with its trading subsidiary, the Cocoa Marketing Company. The reform documents describe the historical role of prefinancing secured against cocoa sales. This is a payment network as much as a physical supply chain. 3 9 18

In a simplified purchase-to-sale cycle, beans are collected, checked, transported and sold. When payment for the sale arrives after the purchasing and handling costs, someone has to bridge the interval. This is a working-capital requirement: cash tied up in inventories and receivables, after allowing for payment terms obtained from suppliers.

The gap between profit and cash appears in other commodity chains too. Our analysis of oil margin calls traces the same principle in a different setting: an economically defensible position can still require immediate liquidity.

The bridge can come from an operator’s own money, a loan or an advance from the final buyer. None is free of constraints. Equity is finite, loans incur interest, and an advance can commit the crop to a particular customer. A contracted future receipt cannot necessarily be used to pay a supplier today.

If funding dries up, the delay can travel all the way back to the grower. Once beans have been delivered but not paid for, the farmer is effectively extending credit to the purchaser. It may be involuntary credit, with no negotiated return and no comfortable repayment schedule. Household expenses still fall due. The official percentage of the cocoa price awarded to producers does not capture that financing burden.

A receivable also needs to be distinguished from cash that is reliably collectible. A buyer may be owed money yet lack a dependable payment date. A bank can therefore decline to fund another collection round even when the anticipated selling price exceeds the cost of the beans. Physical availability and financial availability are separate constraints.

Three prices with different meanings

Three different prices often get collapsed into one. An exchange quote is the price of a standardised contract with its own delivery date and rules. Even the International Cocoa Organization’s daily price indicator is derived from futures contracts in London and New York. It is not a record of transfers to farmers. 8

An export sale on FOB, or free on board, terms is priced for delivery on board at the shipping port. Its economics depend on the particular contract and quality. An exporter’s realised receipts may also combine several generations of sales agreed at different times. The farmgate price, meanwhile, is set in cedis. Subtracting one headline number from another does not reveal the trading margin.

Consider a deliberately hypothetical exchange-rate example. A $5,000 receipt produces GHS50,000 at ten cedis to the dollar, but only GHS45,000 at nine. The cedi has strengthened. That may help an importer while leaving an exporter with fewer cedis to cover an unchanged local bill. A currency hedge or dollar-denominated expenses would alter the exposure.

A meaningful margin calculation therefore needs the actual selling price, the exchange rate used, the costs between collection and shipment, and any earlier contractual commitments. Comparing an exchange quote with an administered farmgate rate can identify a question worth investigating. It cannot, on its own, quantify a loss.

Forward sales require similar care. Agreeing a price before delivery can protect future revenue. If the market subsequently rises, the seller has forgone potential upside; that is not automatically a realised accounting loss. A different problem arises when the crop is smaller than the volume promised. A price hedge cannot supply missing beans.

The important questions are consequently about both price and performance: how much has been committed, what receipts support the borrowing, and what happens if deliveries fall short? Selling forward is neither a universal solution nor proof of reckless management. Its usefulness depends on the risks being hedged and the risks left behind.

Domestic finance changes who bears the risk

The financing plan announced in February envisaged a domestic revolving arrangement: funds raised would purchase cocoa, with sales proceeds repaying the financing within the crop cycle. COCOBOD presented the approach to investors in May. In July it said work was under way on cedi-denominated instruments and the use of pension capital. These were statements about design and preparation. 10 9 11

There is a sound argument for broadening the funding base. Committed money available before collection can reduce dependence on a small set of export counterparties and avoid negotiations conducted under immediate cash pressure. Matching loan maturities to expected receipts also reduces the danger of a repayment falling due before the customer pays.

But a domestic creditor does not make a loan inherently safer. If a local bank or pension fund buys the paper, the exposure ultimately sits with some combination of its shareholders, depositors or retirement savers. A public guarantee may pass part of it to taxpayers. The relevant questions concern contractual rights, security and loss-absorbing capital, not simply the creditor’s nationality.

Currency denomination changes the exposure too. Cedi borrowing removes the direct need to obtain dollars to repay that debt. Yet dollar export receipts can buy fewer cedis if the local currency appreciates. Conversely, borrowing in dollars against dollar sales can provide a partial natural hedge. Neither structure makes currency risk disappear in every circumstance.

The use of proceeds is just as important. Money raised to buy a new crop cannot simultaneously clear all the old liabilities. A revolving fund replenishes itself only if sale collections actually restore its cash. Operating losses can consume the principal; late receipts can force refinancing. Describing the facility as revolving solves neither problem.

As of 6 September, the public sources reviewed do not provide a complete account of a final placement, including cash actually received, interest rate, maturities, guarantees and allocation of proceeds. That is not evidence that no transaction took place. It means the financing outcome cannot be certified from the published intentions alone.

What balance-sheet repair changes

February’s reform plan also envisaged converting debts owed to public institutions into equity and transferring cocoa-road obligations to the state. These proposals concern the balance sheet and the allocation of liabilities. An announcement must not be mistaken for completed implementation. 5

A debt-to-equity conversion can reduce future debt service and improve COCOBOD’s solvency. By itself, it does not deliver new cash. Paying a supplier today still requires a receipt, a cash injection or usable credit. Equally, shifting an obligation to another public entity can remove it from one balance sheet without removing it from the public sector as a whole.

The connection with banks deserves precision rather than alarm. On 25 February, Reuters reported that the buyers’ association estimated members’ bank debts at GHS7–8 billion. That was not an audited banking-system aggregate. Banks emphasised their resilience; the figure alone does not establish a systemic crisis. 12

COCOBOD’s September payment announcement is encouraging for the creditors concerned. Its stated scope is the completion of 2026 DDEP obligations, not repayment of every cocoa-sector liability. The release also contains a small arithmetic inconsistency between its two payment components and stated total. It does not change the order of magnitude, but it is a reason to value independent reconciliation rather than reproduce every total unquestioningly. 2

It would be equally misleading to dismiss all signs of improvement. The IMF’s late-July assessment records substantial progress in macroeconomic stabilisation. In August, the Cocoa Marketing Company announced round-the-clock logistics operations. Faster handling could reduce carrying time and financing costs. There is not, however, an independently measured improvement in cash-collection times attributable to that announcement in the material reviewed. 14 18

The formula behind the new law

The presidency describes the law signed on 26 August as incorporating a 70% share of the world market price for growers and processing at least 50% of beans locally. Preparatory material instead uses the more specific gross-FOB terminology. The full enacted statute and its implementation provisions were not inspected for this article, so a precise link to daily exchange quotations should not be inferred. 1 9

The denominator matters as much as the percentage. Is the reference an achieved average selling price, an anticipated price or a basket of contracts? Which exchange rate applies, and when is the producer rate reset? A transparent formula should let growers understand their pay and financiers anticipate procurement costs. The headline share cannot answer those questions by itself.

Local processing can add value, but it also needs working capital. The processor has to fund beans, conversion and finished-product inventories until buyers pay. A factory does not remove the payment cycle. If it settles later than the export customer it replaces, the financing requirement can increase. Evaluating the change requires actual margins and payment terms, not just a processing target.

Nor has the reform received uniformly enthusiastic support. On 13 August, before assent, Reuters reported concerns from farmer representatives about consultation and restrictions on land use. Their objections should not be treated as the view of every grower, but they complicate a claim of universal endorsement. 13

The global balance is a moving estimate

The world-market data come with a separate warning. In its August bulletin, posted on 31 August, the ICCO temporarily withheld production and grindings estimates for 2025/26. The 37,000-tonne surplus it did report refers to 2024/25. It is not a forecast for the current crop year. 17

Sources: ICCO February, May and August 2026 bulletins. Three estimates of one past crop year, not a monthly series. Surplus = crop net of weight loss − grindings. The August bulletin withholds 2025/26 estimates.

Data and notes

Bulletin	Posted	Surplus (t)
February 2026 bulletin	2026-03-02	75,000
May 2026 bulletin	2026-05-29	48,000
August 2026 bulletin	2026-08-31	37,000

For that same past season, the estimated surplus had been 75,000 tonnes in the February bulletin and 48,000 in May. These are successive estimates of one annual balance, not three monthly harvest observations. The surplus definition subtracts grindings from the crop after adjusting gross production for weight loss. It is not simply gross production minus processing. 15 16

A global balance can help assess aggregate supply and use. It cannot reveal the date on which a particular licensed buyer will be paid. A better-supplied world market can coexist with a local funding shortage. Conversely, stronger cocoa prices will not automatically repair a payment network if future sales are already committed or receipts must first service earlier liabilities.

One hundred tonnes, and the cost of waiting

A small model helps separate these mechanisms. Take an entirely hypothetical 100-tonne lot. Beans cost GHS40,000 per tonne and other upfront expenses GHS3,000. Total expenditure is GHS4.3 million. Selling at $5,000 a tonne, converted at ten cedis per dollar, would produce GHS5 million. The difference before financing is GHS700,000.

With GHS300,000 of the operator’s own cash and no customer advance, required borrowing is GHS4 million. At a fictional annual simple interest rate of 24%, payment after 90 days produces about GHS236,712 of interest, leaving a lot-level result of GHS463,288 before costs excluded from the model. This is an aggregated purchase-and-resale example, not the actual margin of a Ghanaian LBC.

After 180 days, the result falls to roughly GHS226,575. At 270 days it turns slightly negative. Nothing has changed except time. Under these assumptions, interest consumes the initial margin at about 266 days. A liquidity delay has eroded the profitability that the borrowing was meant to finance.

Source: l0g calculations using the included model. Upfront cost GHS4.3m; sales GHS5m; own cash GHS0.3m; loan GHS4m; fictional 24% annual simple interest; no advance. Prices and FX stay constant. Interest = loan × 24% × days / 365. No refinancing risk modelled.

Data and notes

Delay (days)	Interest (GHS)	Result (GHS)
0	0	700,000
30	78,904	621,096
60	157,808	542,192
90	236,712	463,288
120	315,616	384,384
150	394,521	305,479
180	473,425	226,575
210	552,329	147,671
240	631,233	68,767
270	710,137	-10,137
300	789,041	-89,041
330	867,945	-167,945
360	946,849	-246,849

Now suppose the bank commits only GHS3 million against the GHS4 million required. The conditional 90-day margin remains positive, but the purchase has a GHS1 million funding gap. It cannot proceed as entered. At the other extreme, providing more credit to an already loss-making transaction may merely postpone the recognition of the loss.

The calculator lets the reader vary prices, exchange rates, available credit, a buyer’s advance and the payment delay. It reports the initial funding requirement separately from the lot’s result, then checks cash remaining after debt repayment. An advance reduces borrowing; it is not counted as a second sale. Repaying principal is not an expense in the profit calculation, whereas interest is.

The simplifications are explicit: one lot, one final collection, no physical losses, tax, insurance, currency hedges or customer default. Credit is assumed to remain available until payment, so rollover risk is not priced. This is a way to understand transmission, not a valuation of COCOBOD’s accounts or an investment assessment of a cocoa bond.

The decisive evidence will be payment records

The reform can work. The most convincing route is straightforward: cash genuinely raised before collection, controlled use of proceeds, timely payment for beans, collected sales and loans repaid without consuming the next crop’s funding. None of that requires a promise that world prices will remain favourable.

The evidence needed is equally practical: dated arrears and payment delays, net funding actually received, and a reconciliation of volumes sold, achieved prices, exchange rates and cash collected. Annual accounts matter, but they do not replace information on how money moves through a seasonal business.

The available record establishes a financing crisis early in the year, a new legal framework and announced payments to certain creditors. It does not yet certify that every part of the chain has enough usable cash. Until that can be shown, the question is not only what Ghana’s cocoa is worth. It is who finances the wait of the people who grow it.

Sources and reading notes

Observed figures are kept separate from calculator assumptions. COCOBOD statements represent the issuer’s position, not an independent audit. Information cut-off: 6 September 2026.

[1] Présidence de la République du Ghana · 2026-08-26
Mahama signs 10 bills into law to transform economy, modernise justice, and empower citizens.
Assent reported on 26 August. Government summary; full enacted statutory text not inspected.

[2] COCOBOD · 2026-09-01
COCOBOD settles GH¢2.3 billion, completes 2026 DDEP bond obligations
Issuer statement, not an independent audit. Scope: 2026 DDEP payments. The published total exceeds the sum of its two components by GHS54,000.

[3] Reuters · 2026-03-19
Despite price cut, Ghana cocoa buyers lack funds to buy beans from farmers, sources say
March reporting, not a September arrears census. The official price cut was on 12 February, despite an inconsistent relative-date phrase in the report.

[4] Ministry of Finance, Ghana · 2025-10-02
Review of Producer Price for the 2025/26 Cocoa Season
Prices in GHS per tonne: initial announcement on 4 August; revised rate effective 3 October 2025.

[5] Ministry of Finance, Ghana · 2026-02-12
Press briefing on cocoa: financial viability and long-term sustainability
PDF dated 12 February; landing page posted on 13 February. Price: p. 7; restructuring: pp. 5–6. February proposals are not evidence of execution.

[6] COCOBOD · 2026-06-12
Government maintains producer price of cocoa for the 2025/26 light crop season amidst global price decline
Official per-tonne rate for the light crop opening on 18 June 2026. Not a 2026/27 price announcement.

[7] International Monetary Fund · 2026-07
Ghana: 2026 Article IV Consultation and Sixth Review, Country Report 26/212
Section J and Annex VI. Financial diagnosis; prospective legislative timetable superseded by 26 August assent.

[8] International Cocoa Organization · no displayed date
Statistics: ICCO daily price methodology
Futures-based indicator methodology. No current market quote used in this article.

[9] COCOBOD · 2026-05-08
COCOBOD signals new funding model for 2026/27 cocoa season at 2026 ACFIF conference
Funding history and proposed new arrangements.

[10] Ghana News Agency · 2026-02-12
Gov’t unveils domestic bond financing scheme for cocoa purchases
Report on the revolving-fund proposal; state agency, not independent confirmation of a completed placement.

[11] COCOBOD · 2026-07-22
We will achieve efficiency with the new funding model: Chief Executive
Pension capital and cedi instruments contemplated; an expected August funding date does not establish completion.

[12] Reuters · 2026-02-25
Ghana’s cocoa buyers owe banks up to $750 million, raising fresh liquidity risks
Bank-debt range attributed to the buyers’ association, not an audited consolidated banking statement.

[13] Reuters · 2026-08-13
Ghana’s cocoa farmers push back against new law on land use
Concerns expressed before assent; not evidence of unanimous farmer opposition.

[14] International Monetary Fund · 2026-07-27
IMF Executive Board completes Ghana’s sixth ECF review, concludes 2026 Article IV and approves a 36-month PCI
Macroeconomic counterpoint: stabilisation progress, not certification of individual cocoa-sector payments.

[15] International Cocoa Organization · 2026-03-02
February 2026 Quarterly Bulletin of Cocoa Statistics
Release dated 27 February, posted on 2 March. Estimate of the 2024/25 crop year.

[16] International Cocoa Organization · 2026-05-29
May 2026 Quarterly Bulletin of Cocoa Statistics
2024/25 crop year; surplus equals crop net of weight loss less grindings.

[17] International Cocoa Organization · 2026-08-31
August 2026 Quarterly Bulletin of Cocoa Statistics
Posted 31 August, information as of early August; inconsistent “29 May” dateline. 2025/26 data withheld; quoted surplus relates to 2024/25.

[18] COCOBOD / Cocoa Marketing Company · 2026-08-22
CMC goes 24/7 to drive Ghana’s cocoa export momentum
Operational announcement; no independent measure of the resulting acceleration in cash receipts.