The GH¢29.1 billion accumulated losses recorded between 2022 and 2025 under Ghana’s domestic gold purchase operations should not automatically be regarded as losses caused by the Bank of Ghana (BoG) merely because the central bank carried the financial impact on its balance sheet.
The evidence presented in the assessment of the Domestic Gold Purchase Programme (DGPP) points instead to a more complex picture: the commercially generated components of the losses were substantially connected to the gold aggregation and trading activities conducted through the Ghana Gold Board (GoldBod), while a separate portion arose from the exchange-rate treatment applied by the BoG.
That distinction is fundamental to accountability. The institution that records a loss is not necessarily the institution that caused it. In the case of the artisanal and small-scale mining (ASM) gold component of the programme, GoldBod was at the centre of the commercial chain.
It purchased gold, interacted with producers and aggregators, applied or determined service and while paying assay charges, and facilitated sales to off-takers.
The International Monetary Fund (IMF) has identified service and assay fees paid to GoldBod and discounts on gold sold to off-takers among the sources of the losses.
The appropriate conclusion, therefore, is not that every cedi of the GH¢29.1 billion was generated by GoldBod. Rather, the losses should be disaggregated according to their source.
The commercial losses arising from GoldBod’s operations should follow those operations and be reflected on GoldBod’s balance sheet, while the exchange-rate component associated with the BoG’s accounting treatment should be separately identified and attributed to the central bank.
Losses surge to GH¢29.1bn
The scale of the losses makes the question of responsibility particularly important.
Losses under the programme increased from GH¢74.44 million in 2022 to GH¢1.37 billion in 2023, climbed to GH¢5.66 billion in 2024 and then surged dramatically to GH¢22 billion in 2025. Together, the four years produced accumulated losses of approximately GH¢29.1 billion.
The trajectory is striking. The GH¢74.44 million recorded in 2022 increased to GH¢1.37 billion the following year.
By 2024, losses had risen more than fourfold to GH¢5.66 billion, before exploding to GH¢22 billion in 2025 alone.
The 2025 figure therefore represented the overwhelming majority of the four-year cumulative loss.
These figures cannot reasonably be dismissed as a minor accounting variance.
They raise serious questions about the pricing of domestically purchased gold, the charges imposed along the aggregation chain, discounts granted to off-takers, foreign-exchange arrangements and, crucially, the institution that should bear the financial consequences of decisions taken under the programme.
At the heart of the matter is a straightforward accountability principle: an institution that undertakes a commercial activity and makes decisions that generate a financial cost should not automatically be insulated from that cost simply because another state institution provided the financing or eventually recorded the loss.
DGPP was not one transaction
Understanding the dispute requires a closer look at the architecture of the DGPP.
The programme was a government initiative implemented through the BoG and the Precious Minerals Marketing Company (PMMC), which subsequently transitioned into GoldBod.
It was conceived primarily as a macroeconomic intervention rather than a conventional gold-trading operation designed to generate speculative profits.
Its objectives included strengthening Ghana’s foreign-exchange reserves, improving external buffers, reducing pressure on the cedi and supporting broader macroeconomic and exchange-rate stability.
The programme operated through different gold channels.
Under one component, the BoG purchased refined gold from mining companies at London Bullion Market Association (LBMA)-certified refineries.
The refined gold was subsequently placed with international bullion banks as part of the central bank’s reserve-management operations.
The assessment does not identify this channel as the source of the disputed GH¢29.1 billion losses.
A second channel involved the purchase of doré gold from approved aggregators for shipment to LBMA-certified refineries. Selected quantities were refined into monetary gold and, where appropriate, placed with international bullion banks as part of Ghana’s reserve portfolio.
This channel must also be distinguished from the component at the centre of the losses.
The critical channel was the purchase of doré gold from artisanal and small-scale miners through GoldBod.
The gold could be exported in unrefined form to off-takers or gold buyers, generating foreign exchange for market intermediation and reserve-management purposes.
It is this part of the programme that provides the strongest basis for examining GoldBod’s responsibility for the commercial losses.
GoldBod at centre of gold trade
The mechanics of the ASM gold operation provide the clearest explanation of why the losses should not automatically be described as BoG losses.
Under the arrangement, the central bank advanced funds to PMMC and subsequently GoldBod in cedis to finance the purchase of gold from the domestic market.
GoldBod acquired the gold from the domestic supply chain, after which the gold was sold to off-takers and converted into United States dollars.
The dollars were deposited with the BoG and became part of the foreign-exchange resources available to the central bank for reserve management and economic stabilisation.
Put simply, the BoG supplied the cedi financing and received the resulting foreign exchange, while GoldBod occupied the centre of the domestic gold aggregation and trading operation.
That distinction matters. The BoG’s primary role was linked to financing, receiving foreign exchange and managing reserves. GoldBod’s role was much closer to the commercial transaction itself.
It operated within the domestic gold market and interacted with gold producers, aggregators and off-takers.
Its activities therefore directly affected important elements of the cost structure of the gold trade.
This does not mean that every component of the GH¢29.1 billion originated from GoldBod.
It means that the losses must be traced to the transactions and decisions that produced them instead of automatically being assigned to the central bank simply because the BoG’s balance sheet carried the financial impact.
IMF breakdown points to commercial losses
The IMF’s explanation of the losses is particularly important because it provides the basis for separating commercial losses from the accounting component.
According to the assessment, losses on gold trades were a combination of service and assay fees paid to GoldBod, discounts on gold sold to off-takers and, most importantly, exchange-rate losses arising from the spread between the forex bureau rate paid to purchase gold and the cedi reference rate used by the BoG for accounting purposes.
The first two components have a direct connection with the gold-trading operation.
Service and assay fees are costs associated with aggregating, testing and preparing gold for sale.
Discounts to off-takers similarly arise from the commercial disposal of the commodity.
The source material indicates that GoldBod determined the service and assay fees and the discounts applied to gold sold to off-takers.
This creates a straightforward accountability question.
If an institution is responsible for a commercial transaction, determines charges associated with that transaction and manages the sale of the commodity, why should another institution automatically absorb the entire financial consequence?
The BoG’s conventional function is monetary and financial stability, including reserve management.
It is not ordinarily the commercial operator determining fees within a domestic commodity aggregation chain.
Consequently, the commercially generated component of the losses should logically follow the institution responsible for the underlying activity.
The BoG has a distinct exchange-rate responsibility
There is, however, an important qualification.
The exchange-rate component requires different treatment. GoldBod paid the forex bureau rate when purchasing the gold, while the BoG used its cedi reference rate to account for the transactions.
The difference between those rates generated an exchange-rate loss.
That component has a clearer connection with the central bank’s accounting framework.
If the BoG’s reference rate differed from the actual rate at which the gold purchases were funded, the accounting treatment created a variance between the cedi value assigned to the transaction and the actual amount paid.
It would therefore be inaccurate to argue that every cedi of the GH¢29.1 billion was caused by GoldBod.
The exchange-rate component should be separately identified and attributed according to the decision or accounting mechanism that created it.
But that qualification strengthens rather than weakens the case for proper attribution.
It means the GH¢29.1 billion should be subjected to component-by-component accounting rather than being treated wholesale as a BoG loss.
Who made the commercial decisions?
The crucial question is not simply who provided the money.
It is who made the decisions that determined the commercial cost of acquiring and disposing of the gold.
GoldBod’s role in the ASM gold chain was central.
It purchased gold from the domestic market using funds advanced under the programme, participated in the aggregation process, determined service and assay fees, facilitated sales to off-takers and applied discounts.
Those activities directly affected the margin between the amount spent to acquire the gold and the value ultimately realised from its sale.
If fees increased the cost of the gold or discounts reduced proceeds from its disposal, the resulting financial impact arose from the commercial structure of the transaction.
That is fundamentally different from the BoG’s role as the institution managing the foreign exchange received from the sale.
The distinction can be reduced to one principle: carrying a loss is not necessarily causing a loss.
An institution can carry a loss because of a government policy decision without being the institution whose commercial activities generated that loss.
Why GoldBod should carry the commercial losses
The argument for placing commercially generated losses on GoldBod’s balance sheet rests on accountability.
An institution should be financially accountable for activities it controls. GoldBod was not a passive beneficiary of the DGPP.
It operated the critical domestic ASM gold-purchasing channel and performed functions including acquiring gold, dealing with the domestic supply chain, applying service and assay charges and selling gold to off-takers.
The IMF’s own breakdown identifies service and assay fees paid to GoldBod and discounts on gold sold to off-takers as sources of the losses.
Those are commercial variables, not merely consequences of monetary policy. They form part of the mechanics of the gold transaction.
If those costs are absorbed entirely by the BoG, the central bank’s financial statements may show a substantial loss without adequately revealing the institution or commercial activity that generated the underlying cost.
That weakens accountability and makes it harder to assess whether the gold-purchasing mechanism was efficiently designed.
Recording the commercial loss at GoldBod would not necessarily mean that the government must abandon the programme. Instead, it would allow policymakers to determine whether GoldBod’s pricing model, fee structure, discounts and purchasing arrangements are financially sustainable and whether the institution requires restructuring, a subsidy or recapitalisation.
GoldBod’s govt-programme defence
GoldBod’s reported position that it should not carry the losses because the DGPP was a government programme presents a legitimate counterargument.
The institution did not conceive the DGPP as an independent commercial venture.
It operated within a broader government strategy intended to use Ghana’s gold resources to strengthen foreign-exchange reserves and support economic stability.
GoldBod could therefore argue that it was implementing government policy rather than pursuing an independent commercial objective.
But this does not completely resolve the accountability question.
An institution can implement a government programme while remaining responsible for the financial consequences of activities it controls.
If government decides that GoldBod should not bear a particular loss because the transaction was undertaken in the national interest, the state can explicitly compensate or recapitalise the institution.
What should be avoided is the simple transfer of losses from one state balance sheet to another without a transparent explanation of why that transfer is justified.
The public deserves to know whether the losses resulted from policy decisions, operational decisions, pricing decisions, accounting decisions or a combination of these factors.
Why the Ministry of Finance matters
The dispute eventually involved the IMF, GoldBod, the BoG and the Ministry of Finance, with the outcome that losses from the gold programme would be transferred to the Ministry of Finance.
That may resolve an immediate accounting dispute, but it does not resolve the underlying accountability question.
Moving a loss from one state institution to another does not make the loss disappear.
Ultimately, the cost remains a public-sector cost and is therefore borne by the Ghanaian state, taxpayers and citizens. That is why the source of the loss matters.
If the commercial portion was generated by GoldBod’s operations, recording it at GoldBod would provide a clearer picture of the actual cost of those operations.
If government wants the cost to be centrally borne because the programme is a national stabilisation intervention, that decision should be explicit rather than allowing the BoG to become the default balance sheet for losses generated elsewhere in the gold-trading architecture.
G4R was not designed for speculative profit
Another important distinction concerns the Gold for Reserves (G4R) component.
Under G4R, doré gold was acquired primarily to generate foreign exchange and was disposed of within the shortest possible period after purchase.
The gold was not held for price appreciation or trading gains.
The objective was to convert Ghana’s gold resources into foreign exchange and strengthen the country’s reserve position.
That policy objective can nevertheless generate financial losses where the acquisition cost, associated fees, discounts and exchange-rate treatment result in sale proceeds being lower than the cedi value assigned to the acquisition.
But even in that situation, the accounting treatment should identify the source of the loss.
If the loss arises from GoldBod’s aggregation and sale structure, that component should be recognised as a GoldBod commercial loss. If it arises from the BoG’s exchange-rate accounting, it should be separately identified as such.
G4O must not be lumped together
The Gold for Oil (G4O) programme adds another dimension and should not simply be lumped together with G4R.
Under G4O, gold was sold to support foreign-exchange requirements associated with petroleum-product procurement. The financial outcome reflected the combined effect of gold sales and oil-trading activities after directly attributable selling and related costs.
The mechanics were therefore different from those of G4R.
For proper accountability, each component should have its own financial record.
The public should be able to establish how much was lost through G4R, how much through G4O, how much through ASM gold aggregation and how much resulted from exchange-rate accounting.
GH¢29.1bn requires forensic disaggregation
The most compelling response to the controversy would be a complete breakdown of the GH¢29.1 billion.
The public should know how much of the GH¢74.44 million loss in 2022 arose from service and assay fees, how much came from discounts to off-takers and how much resulted from exchange-rate differences.
The same analysis should be conducted for the GH¢1.37 billion recorded in 2023, the GH¢5.66 billion in 2024 and the GH¢22 billion in 2025.
Such a breakdown would establish whether the overwhelming portion of the losses was generated by GoldBod’s commercial operations or by the BoG’s exchange-rate accounting.
Without that disaggregation, it would be too sweeping to claim that every cedi of the GH¢29.1 billion was caused by GoldBod. But the available explanation provides a strong basis for insisting that commercial losses should be attributed to the institution operating the commercial gold chain rather than automatically classified as central-bank losses.
Accountability must follow the transaction
The controversy exposes a broader weakness in the institutional architecture of Ghana’s gold-for-reserves strategy.
The objective of using Ghana’s gold resources to strengthen foreign-exchange reserves, reduce vulnerability to external shocks and support the cedi is legitimate.
The DGPP was designed to improve reserve buffers, support exchange-rate stability and strengthen confidence in the economy.
But macroeconomic benefits do not eliminate the need to account for the cost of achieving them.
A programme can strengthen Ghana’s foreign-exchange position while simultaneously generating substantial financial losses.
The proper response is therefore not necessarily to abandon the policy, but to establish precisely where the losses originate and ensure that the institutional architecture reflects those realities.
The strongest case for GoldBod carrying the commercially generated portion is not that GoldBod independently created the entire DGPP or that the BoG played no role.
Rather, it is that GoldBod occupied the operational centre of the specific gold-trading channel associated with the disputed losses.
The BoG provided funding and managed the foreign exchange generated by the transactions.
GoldBod operated the domestic gold-purchase and sales mechanism.
The IMF’s explanation specifically identifies GoldBod-linked service and assay fees and discounts to off-takers among the sources of losses.
The appropriate conclusion is therefore differentiated responsibility.
Commercial losses should follow GoldBod’s operations, while the exchange-rate component should be separately attributed to the BoG’s accounting treatment.
A public-interest issue
Ultimately, the GH¢29.1 billion controversy is a public-interest issue.
Ghana’s gold belongs to the nation, and the institutions established to manage that resource must operate with the highest standards of transparency and accountability.
A system in which one institution makes commercial decisions, another finances the transactions, a third ultimately absorbs the losses and the Ministry of Finance later assumes responsibility creates the risk of blurred accountability.
Taxpayers should not have to settle for knowing which institution recorded the loss.
They should know why the loss occurred, who made the relevant decisions, whether those decisions were reasonable, whether the pricing arrangements were efficient and whether the institutional structure created incentives for financial discipline.
The GH¢29.1 billion episode therefore provides an opportunity to redesign the financial architecture of the domestic gold-purchase strategy.
Every transaction should have a clearly identifiable financial owner.
Every service charge, assay fee and discount should be transparently recorded.
Every exchange-rate difference should be separately disclosed.
G4R and G4O should have distinct accounts, while the commercial gold-aggregation function should be separated from the central bank’s reserve-management function.
Where government deliberately chooses to subsidise a programme for macroeconomic reasons, that subsidy should also be explicitly recognised rather than hidden within the financial statements of another state institution.
Such an arrangement would protect the integrity of both GoldBod and the BoG while making it easier to determine whether the country’s gold-for-reserves strategy delivers sufficient macroeconomic benefits to justify its financial cost.
The real question: who generated the loss?
The GH¢29.1 billion should therefore not be reduced to a political argument over whether the BoG or GoldBod “lost” the money.
The more important question is: who generated each component of the loss?
The programme’s structure indicates that commercial components associated with service and assay fees and discounts to off-takers are closely linked to GoldBod’s gold-trading operations.
The exchange-rate component, by contrast, is linked to the difference between the forex bureau rate used for gold purchases and the BoG’s cedi reference rate used in accounting.
That distinction provides the strongest basis for placing the commercially generated portion of the GH¢29.1 billion on GoldBod’s balance sheet rather than automatically charging it to the BoG.
The BoG’s absorption of the losses reflects the policy architecture under which the programme operated and its mandate to accumulate foreign-exchange reserves and support exchange-rate stability. But carrying the loss does not necessarily mean causing the loss.
GoldBod must therefore be accountable for the commercial consequences of transactions it controls; the BoG must be accountable for exchange-rate and accounting decisions within its mandate; and the Ministry of Finance must be accountable for any decision to transfer programme losses to the public finances.
The central lesson is clear: the institution that carries a loss should not automatically be assumed to be the institution that caused it.
Ghana must follow the transaction, identify the decision that generated the cost and assign responsibility accordingly.
On the evidence presented, the commercially generated losses from GoldBod’s ASM gold-purchasing and trading operations should be recognised on GoldBod’s books, while the clearly identifiable exchange-rate accounting component should remain attributable to the BoG.
Only such a component-by-component approach can give Ghanaians a truthful picture of the cost of the country’s gold-for-reserves strategy and prevent the central bank from becoming the default balance sheet for losses generated elsewhere in the gold-trading architecture.
