IMF, GoldBod and BoG’s $1.7bn Loss: What the Report Actually Says Amid Afenyo-Markin–Sammy Gyamfi Dispute

The political controversy surrounding Ghana’s domestic gold programme has intensified following conflicting interpretations of reports by the International Monetary Fund (IMF).

At the centre of the dispute are three major questions: Did Ghana actually lose billions of dollars through its gold programme? Did GoldBod cause the losses? And are the losses real economic losses or largely accounting and exchange-rate adjustments?

A review by NsemGH of the IMF documents and statements from Minority Leader Alexander Afenyo-Markin and GoldBod Chief Executive Officer Sammy Gyamfi shows that the truth is more complex than either side’s political messaging suggests.

What the IMF actually reported

In a Selected Issues Paper published in August 2026, the IMF examined the Bank of Ghana’s Domestic Gold Purchase Programme, which was introduced in 2021.

The programme allowed the Bank of Ghana to purchase gold locally to build Ghana’s international reserves, support the Gold-for-Oil initiative and provide foreign exchange to the domestic market.

According to the IMF, the programme helped Ghana to formalise artisanal and small-scale gold exports, rebuild its foreign reserves and stabilise the foreign exchange market during the country’s economic crisis.

However, the IMF also found that the programme generated substantial losses for the Bank of Ghana and exposed the central bank to what it described as “quasi-fiscal risks.”

The Fund estimated that losses under the programme amounted to almost US$400 million in 2024, equivalent to approximately 0.5 per cent of Ghana’s Gross Domestic Product.

Following a significant expansion of domestic gold purchases in 2025, the losses associated specifically with gold transactions reportedly exceeded US$1.7 billion.

A separate IMF programme review placed the wider losses associated with the Domestic Gold Purchase Programme at approximately US$2.2 billion, or about 1.5 per cent of GDP, in 2025.

The difference between the US$1.7 billion and US$2.2 billion figures appears to be one of scope. The US$1.7 billion figure relates mainly to losses from gold-trading operations, while the larger figure reflects the broader financial impact of the programme on the Bank of Ghana.

Where did the earlier US$214 million figure come from?

Before the latest IMF assessment, public debate had focused on a US$214 million loss.

The IMF’s December 2025 review reported that, by the end of September 2025, losses from the artisanal and small-scale doré-gold component of the Gold-for-Reserves programme had reached US$214 million.

That was an interim figure covering only part of 2025 and a narrower component of the programme. It should therefore not be presented as the programme’s final loss for the entire year.

The later figures of US$1.7 billion and US$2.2 billion were based on a broader assessment of the expanded programme and its impact on the Bank of Ghana.

How were the losses generated?

The IMF did not suggest that billions of dollars in physical cash simply disappeared. Its assessment identified several components behind the financial losses:

• Service and assay fees paid in connection with purchasing and testing gold.
• Discounts applied when the gold was sold to international off-takers.
• The difference between the price at which gold was purchased and the price at which it was sold.
• Exchange-rate differences between the forex-bureau rates used to purchase gold locally and the Bank of Ghana’s official cedi reference rate used for accounting.
• Valuation effects resulting from changes in the cedi and the way gold and foreign-currency transactions were recorded.

The IMF indicated that the losses were equivalent to approximately 17 per cent of the value of the doré gold sold by the Bank of Ghana in 2025.

It also observed that prices paid for Ghana’s artisanal and small-scale gold were among the highest in the region.

The Fund nevertheless acknowledged that part of the headline amount represented accounting and exchange-rate valuation effects rather than a direct economic cost of the same magnitude.

This distinction is important. An accounting loss remains a legitimate concern because it weakens the Bank of Ghana’s balance sheet, but it is not necessarily evidence that an equivalent amount of money was stolen, diverted or physically lost.

What Afenyo-Markin is alleging

Minority Leader Alexander Afenyo-Markin has argued that the IMF’s findings confirm a loss of approximately US$1.7 billion—estimated at about GH¢22 billion—through the domestic gold programme.

He maintains that GoldBod cannot completely distance itself from the loss because it played a central role in purchasing, aggregating and assaying gold for the Bank of Ghana while receiving service and assay fees.

The Minority Leader has questioned why GoldBod should earn fees from the transactions while the Bank of Ghana bears the losses and associated financial risks.

He is demanding details about the prices paid for gold, the method used to determine those prices, the selection of off-takers, the discounts applied during sales and the fees received by GoldBod.

These are legitimate accountability questions, particularly because public funds and the Bank of Ghana’s balance sheet were involved.

However, describing the entire US$1.7 billion as a “GoldBod loss” goes beyond the precise wording of the IMF report.

The IMF attributed the losses to the Bank of Ghana’s Domestic Gold Purchase Programme. It did not expressly conclude that GoldBod alone caused or incurred the entire amount.

Sammy Gyamfi’s defence

Sammy Gyamfi has rejected claims that GoldBod itself lost US$1.7 billion or that the IMF blamed the institution for the Bank of Ghana’s losses.

He has challenged Afenyo-Markin to identify any section of the IMF reports that directly states that GoldBod caused the US$1.7 billion loss.

On this narrow point, Gyamfi is correct: the IMF described the losses as losses of the Bank of Ghana under the Domestic Gold Purchase Programme. It did not label the figure as GoldBod’s corporate loss.

Gyamfi also argues that GoldBod’s role in 2025 was limited to purchasing and aggregating gold as an agent for the Bank of Ghana.

According to him, GoldBod did not determine the selling price, select the off-takers or sign the agreements under which the Bank of Ghana sold the gold.

He further points out that the programme recorded almost US$400 million in losses in 2024, before GoldBod was formally established. GoldBod was created in 2025 to replace and expand functions previously performed by the Precious Minerals Marketing Company and other buying arrangements.

Gyamfi says GoldBod received a 0.5 per cent service fee and a 0.258 per cent assay fee for its work. He maintains that these were legitimate, legally approved fees and did not make the institution responsible for all the losses recorded by the Bank of Ghana.

He has also cited GoldBod’s financial statements, saying the institution recorded an operational surplus of GH¢907 million and an overall surplus exceeding GH¢5.4 billion in 2025.

However, GoldBod recording a surplus in its own accounts does not automatically cancel the losses recorded by the Bank of Ghana. The two institutions have separate financial statements and responsibilities.

It is possible for GoldBod to earn fees and record a surplus while the Bank of Ghana records losses from financing, exchange-rate movements and the subsequent sale of the gold.

Who is telling the truth?

Both sides are relying on parts of the available evidence, but their conclusions require qualification.

Afenyo-Markin is correct that the IMF reported very substantial losses under the domestic gold programme and that the financial consequences ultimately affected the Bank of Ghana and the public sector.

He is also justified in demanding transparency regarding GoldBod’s fees, purchase prices, off-takers, discounts and risk-sharing arrangements.

However, the IMF did not describe the entire US$1.7 billion as a loss made solely by GoldBod. Presenting it simply as a “GoldBod loss” is therefore misleading.

Sammy Gyamfi is correct that the IMF formally recorded the losses against the Bank of Ghana’s programme and did not directly accuse GoldBod of causing the entire amount.

He is also correct that some of the losses predated GoldBod and that exchange-rate and accounting valuation effects contributed substantially to the figures.

Nevertheless, saying that GoldBod had nothing to do with the losses would also be incomplete. GoldBod participated in the gold-purchasing chain, received service and assay fees and was one of the institutions involved in implementing the programme during 2025.

The IMF specifically identified service and assay fees—including payments associated with GoldBod—as one component of the programme’s costs. This does not make GoldBod responsible for the entire loss, but it means its activities cannot be considered completely unrelated.

GoldBod operations transferred from BoG

The IMF reported that the government, the Bank of Ghana and GoldBod signed a memorandum of understanding in July 2026 to formalise the transfer of the Domestic Gold Purchase Programme’s activities from the central bank to GoldBod.

The objective is to remove gold-trading and other quasi-fiscal risks from the Bank of Ghana’s balance sheet.

This development supports Gyamfi’s argument that the programme was previously operated and financed primarily through the Bank of Ghana. It also strengthens the IMF’s position that a central bank should not continue undertaking commercial or quasi-fiscal gold-trading activities that can weaken its financial position.

The transfer, however, makes strong oversight of GoldBod even more important because the institution will now assume greater responsibility for domestic gold trading.

External audit expected to provide clearer answers

The IMF disclosed that an independent external audit of the Domestic Gold Purchase Programme, covering its operations from its inception in 2021, was underway.

That audit is expected to establish the exact economic and accounting losses, the amounts paid in fees, the pricing and off-taker arrangements, and the contribution of exchange-rate movements to the Bank of Ghana’s losses.

Until the detailed audit is published, claims of corruption, theft or deliberate financial wrongdoing should be treated as allegations rather than established facts.

NsemGH’s conclusion

The IMF did report serious losses under the Bank of Ghana’s Domestic Gold Purchase Programme. Those losses cannot be dismissed simply because the programme contributed to reserve accumulation and exchange-rate stability.

However, the IMF did not say that GoldBod alone lost or misappropriated US$1.7 billion.

The most accurate description is that the Bank of Ghana recorded substantial financial and accounting losses through the Domestic Gold Purchase Programme, with purchasing fees, gold-sale discounts, high domestic purchase prices and exchange-rate valuation differences contributing to the outcome.

GoldBod was part of the purchasing structure and earned fees, but the evidence currently available does not support attributing the entire loss exclusively to GoldBod or Sammy Gyamfi.

The controversy therefore requires greater transparency—not political exaggeration. The government, Bank of Ghana and GoldBod should publish the full transaction records, audited financial statements, off-taker arrangements and the independent audit when completed.

Only that disclosure will allow Ghanaians to determine the true economic cost of the programme and whether the losses arose from policy design, poor commercial decisions, accounting treatment, negligence or wrongdoing.

Sources: IMF Selected Issues Paper on the Bank of Ghana’s Domestic Gold Purchase Programme⁠, IMF Fifth Review report⁠, Afenyo-Markin’s position⁠, and Sammy Gyamfi’s response⁠.