Sika Kɔkɔɔ Without Tears

We are still Gold Coast. Gold is now responsible for more than half of our exports, compared with about one-fifth in 2021. So when we talk about gold today, we are talking about one of the most important drivers of Ghana’s external economy.

The gold comes from two main sources. Large-Scale Gold Mining (LSGM), the big mining companies, and Artisanal and Small-Scale Gold Mining (ASGM) or the much more fragmented small-scale sector. In 2025, ASGM production actually overtook large-scale mining production for the first time.

For years, a lot of small-scale gold was leaving Ghana unofficially. The IMF estimates that between 2019 and 2024, about 229 tonnes of gold worth $11.4 billion were missing from Ghana’s official trade statistics. That does not necessarily mean all $11.4 billion was “stolen.” It means enormous quantities of gold were apparently leaving outside the recorded official export system. That matters because Ghana loses visibility, foreign exchange and potentially tax revenue when gold disappears through unofficial channels.

Enter the Domestic Gold Purchase Programme (DGPP). Bank of Ghana introduced the DGPP in 2021 at a time when we were facing falling reserves, exchange-rate pressure and increasingly difficult access to foreign financing. The basic idea was simple. We produce gold. Why not use some of that gold to strengthen our foreign reserves and meet our foreign-exchange needs?

ADVERTISEMENT

But “DGPP” is not one single transaction. DGPP is an umbrella covering different gold operations. (a) Gold-for-Reserves (G4R) or gold used to build Ghana’s foreign reserves. And within G4R, we must distinguish between bullion and doré gold. (b) Gold-for-Oil (G4O) or gold or proceeds from gold used to finance petroleum imports. G4O ended in May 2025. These distinctions matter because their economics are very different.

Bullion is not the same thing as doré. Bullion is refined gold. Under the programme, bullion was purchased mainly from large-scale mining companies, generally at market prices, refined through LBMA-certified refineries and held abroad as reserves. Doré is semi-refined gold, much of it originating from the small-scale mining sector. Getting that gold from the miner into dollars in Ghana’s reserves requires several steps. And every step can have a cost.

Enter GoldBod. GoldBod was created in 2025 and became central to sourcing small-scale gold. It sources gold through aggregators and traders. Under the arrangement described by the IMF, GoldBod essentially acted as the sourcing/broker institution while Bank of Ghana carried the gold purchases and associated financial exposure under the DGPP. A surplus reported by GoldBod and a loss reported by Bank of Ghana are therefore not necessarily contradictory. They can be reporting different parts of the same chain.

The scale became enormous. In 2025, BoG purchased and exported approximately 104 tonnes of ASGM gold worth $10.9 billion. That was essentially all officially recorded small-scale gold production. BoG also handled approximately $1.26 billion of bullion.

DGPP achieved some remarkable things. Gold-related inflows into BoG rose to approximately $12.7 billion in 2025. Gross international reserves reached $11.9 billion, about four months of imports and significantly above programme targets. The programme also gave BoG enormous capacity to supply foreign exchange to the economy. BoG sold $10.6 billion of FX in 2025, compared with just $1 billion in 2023.

The IMF says these sales improved FX liquidity and coincided with the cedi appreciating 41% against the dollar in 2025. So the DGPP was not a programme that produced nothing. It played an important role in Ghana’s macroeconomic stabilization.

But there is an expensive side to the story. According to the IMF, DGPP operations generated losses of almost $400 million in 2024. Then, as the programme expanded dramatically in 2025, so did the losses, reaching over $1.7 BILLION (about 1.5% of GDP).

The IMF says the 2025 losses were almost entirely related to G4R doré purchases. So saying simply “We sold gold and lost $1.7 billion” misunderstands the economics or distorts the story.

When my cousin, the Mpraeso koko seller, says that, I can forgive her. But I blame those of us who should explain it so she does not say that.

The bullion story was very different. The IMF reports approximately $1.1 billion in net gains from gold bullion sales in 2025. So Doré operations generated enormous losses. Bullion sales generated substantial net gains. That is why we must stop treating every government gold transaction as if it were the same transaction.

So where did the $1.7 billion loss come from? The IMF identifies three major components.

First: Service and assay fees. GoldBod and others have to aggregate, test, process and facilitate the gold transactions. Those services cost money.

Second: Discounts to off-takers/exporters. BoG did not necessarily receive the full international value when doré gold was sold. The difference contributes to the loss.

Third, and according to the IMF, MOST IMPORTANTLY, exchange-rate differences. This is the part that deserves careful explanation.

Suppose BoG acquires gold requiring $1 billion of foreign-exchange value, but the cedi cost is determined using the Abotsi exchange rate of ₵15 = $1. BoG pays the equivalent of ₵15 billion.

Now suppose the corresponding $1 billion foreign-exchange value is accounted for using BoG’s reference rate of ₵12 = $1. That $1 billion translates into only ₵12 billion.

We now have a ₵3 billion exchange-rate/accounting loss. The important point is the existence of that ₵3 billion accounting loss does not mean ₵3 billion disappeared.

Nor, by itself, does it establish looting, crime or misappropriation.

It tells us that BoG incurred a loss because of the exchange-rate structure of the transaction. And according to the IMF, this exchange-rate spread was the most important source of the DGPP losses.

But does that mean the loss is just accounting hocus-pocus? No. The IMF itself gives us the necessary caution. It says “These accounting losses partly reflect valuation effects rather than economic costs.” PARTLY means some of the loss is economic.

We should therefore NOT automatically interpret the entire $1.7 billion as $1.7 billion in cash that vanished from Ghana. And we certainly cannot jump from:

“BoG recorded a $1.7 billion loss” to: “Somebody stole $1.7 billion.”

Those are entirely different propositions. But neither should we wave the losses away as meaningless accounting. The IMF says the losses weakened BoG’s balance sheet and implied transfers to those receiving FX at the reference rate.

So, my brothers and sisters:

Accounting loss ≠ imaginary loss.

Accounting loss ≠ necessarily equivalent economic cost.

And above all:

LOSS ≠ LOOT.

The real task is to understand the loss, identify who bore the economic cost and who received the corresponding benefit, and determine whether the same policy objectives could have been achieved more cheaply.

This also explains why GoldBod can report a surplus while BoG reports losses. They are different entities performing different functions and keeping different books. If GoldBod receives service or assay fees, those may be income to GoldBod. But those same fees may be costs to Bank of Ghana. One entity’s revenue can literally be another entity’s expense.

So GoldBod surplus ≠ DGPP profit. BoG loss ≠ GoldBod loss.

You must know whose books you are looking at. Separate Entity Concept!

The programme has now changed. From July 1, 2026, DGPP activities were transferred fully to GoldBod. According to the IMF, BoG will no longer carry the programme’s quasi-fiscal financial risks in the same way. The government will assume the programme’s costs transparently. That should make future accountability easier, provided the reporting is sufficiently detailed, as it should.

The good news is that costs appear to be falling. The IMF says DGPP costs fell from approximately 14.5% of gold purchased in 2025 to 11.4% in the first quarter of 2026. Assay fees, service charges and logistical costs were reduced. The troublesome gap between the forex-bureau exchange rate and BoG reference rate also narrowed significantly.

The target is to bring total costs down to about 5%. But even 5% of billions of dollars is serious money.

There is another cost that hardly features in the public debate: sterilization. When BoG creates cedis to purchase gold and accumulates foreign reserves, it may subsequently have to remove excess cedi liquidity from the economy. That costs money too. The IMF warns that Ghana’s plan to build reserves to 15 months of import cover could push open-market-operation costs alone to around 3% of GDP.

So buying gold is not free merely because the gold or its proceeds remain an asset. The financing structure matters.

And gold has another cost that cannot be found on BoG’s balance sheet. The IMF warns that rapid small-scale mining is damaging farmland, cocoa farms, forests, rivers and water supplies, and ultimately threatening Ghana’s long-term agricultural productivity.

A gold programme cannot be called successful merely because it generates dollars while destroying the productive assets that must sustain future generations.

20. So what questions should citizens be asking?

Not the political goldmine: “Who stole the $1.7 billion?”

And not simply: “Didn’t the programme strengthen the cedi?”

Both questions are too crude!

The useful questions are:

How much gold was purchased?

What did we pay for it?

At what exchange rate?

What fees were paid?

To whom?

What discounts were given to off-takers?

How much did the gold ultimately sell for?

How much FX was generated?

What portion went into reserves?

What portion was sold into the FX market?

What were the actual economic costs?

What portion of the reported loss was accounting/valuation?

And who ultimately benefited from the exchange-rate differential?

Those numbers should be capable of reconciliation.

21. The final lesson is that we need to separate things that look similar but are not:

GoldBod ≠ Bank of Ghana.

DGPP ≠ one programme.

G4R ≠ G4O.

Doré ≠ bullion.

GoldBod surplus ≠ DGPP profit.

BoG loss ≠ GoldBod loss.

Accounting loss ≠ cash disappeared.

LOSS ≠ LOOT.

Accounting loss ≠ nothing to worry about.

The interesting question is not whether Ghana should use its gold strategically. Of course it should. The question is whether we can obtain the benefits (formalized exports, foreign exchange, stronger reserves and reduced smuggling) without paying ₵1 when 80 or 85 pesewas would do the job. That is where the serious national conversation should begin.

We explain. You decide.