GPExGhana Policy Exchange← Policy Tracker

Provisional policy-team analysis · 17 August 2026

Gold programme built reserves at substantial cost; IMF and BoG figures remain unreconciled

The evidence supports serious monetary-governance concern—but not the claim that GoldBod itself lost US$1.7 billion.

Executive assessment

The programme’s benefits and its very substantial public cost can coexist.

IMF analysis indicates that the Bank of Ghana-led Domestic Gold Purchase Programme incurred major losses in 2025, concentrated in the Gold-for-Reserves doré channel. GoldBod’s standalone audited accounts, by contrast, report an operational surplus. These statements concern different entities, accounting boundaries and methodologies; they are not contradictory. The central question is whether reserve accumulation, foreign-exchange availability and gold-sector formalisation justified the programme’s complete economic cost and central-bank risk.

Evidence position

What the available evidence supports

Institutional accounts and programme-wide estimates answer different questions. GPEx therefore retains the boundary around each claim.

  1. 01

    The IMF identified US$214 million in Gold-for-Reserves losses on artisanal and small-scale doré transactions through the third quarter of 2025, largely from trading losses and GoldBod off-taker fees.

    Verified
  2. 02

    Later IMF analysis reported losses of more than US$1.7 billion for the expanded 2025 programme, approximately 1.5% of GDP and concentrated in the doré-gold channel.

    Qualified
  3. 03

    The Bank of Ghana’s audited 2025 financial statements separately report a GH¢9.05 billion loss from the Domestic Gold Purchase Programme.

    Audited
  4. 04

    GoldBod’s audited institutional accounts report a GH¢909.71 million operational surplus; they do not report a US$1.7 billion corporate loss.

    Audited
  5. 05

    The programme contributed materially to reserve accumulation, but the available evidence does not establish its consolidated net economic benefit.

    Mixed

Reconciliation

Three figures, three accounting boundaries

None of these figures should be substituted for another without a consolidated reconciliation.

IMF programme estimate>US$1.7bn

A broader estimate of 2025 DGPP losses, mainly in the Gold-for-Reserves doré channel. It includes service and assay charges, off-taker discounts, trading effects and exchange-rate effects; sterilisation costs are excluded.

Bank of Ghana audited accountsGH¢9.05bn

The Bank of Ghana’s separately reported 2025 DGPP loss, including approximately GH¢544 million attributed to Gold-for-Oil.

GoldBod audited accountsGH¢909.71m

GoldBod’s reported operational surplus for 2025 after recording non-tax revenue and institutional expenditure.

Potential benefits

Identifiable macroeconomic contribution

  • International reserve accumulation and greater official gold holdings.
  • Increased formalisation of gold exports and capture of export proceeds through official channels.
  • Additional foreign-exchange availability during a period of wider macroeconomic stabilisation.

Economic costs

Programme and balance-sheet exposure

  • Purchasing premiums, assay and service charges, off-taker fees and discounts.
  • Trading losses, exchange-rate effects and financing costs.
  • Central-bank balance-sheet exposure and other quasi-fiscal costs, including sterilisation costs not captured in the IMF estimate.

Governance assessment

The central issue is programme design and institutional responsibility.

Large-scale quasi-commercial purchasing exposed the central bank to pricing, trading and financing risks beyond conventional monetary operations. The IMF has called for greater transparency, reduced quasi-fiscal activity and budget recognition of relevant costs. Bank of Ghana prefinancing ended from 1 July 2026, moving purchasing risk away from the central bank; this is a potentially important correction, not a settlement of the costs already incurred.

Fact-check boundary

What is not established

These conclusions go beyond, or misstate, the evidence presently available.

  1. 01

    That GoldBod itself incurred a US$1.7 billion corporate loss.

    Not established
  2. 02

    That US$1.7 billion was stolen, misappropriated or caused by corruption.

    Not established
  3. 03

    That the programme produced no macroeconomic benefits.

    Not established
  4. 04

    That GoldBod’s institutional surplus proves that the wider public programme was financially efficient.

    Not established
  5. 05

    That changes in the cedi, inflation or reserve position were caused exclusively by the gold programme.

    Not established
  6. 06

    That the replacement commercial-financing model has demonstrated long-term financial sustainability.

    Not established

Evidence gaps

Questions requiring a consolidated public reconciliation

These questions determine whether Ghana obtained adequate value for the financial cost and how responsibility should be allocated.

  1. 01

    What is the complete decomposition of the IMF’s reported US$1.7 billion estimate, including realised cash losses and accounting or valuation effects?

    Open
  2. 02

    Why does the IMF estimate differ from the GH¢9.05 billion loss in the Bank of Ghana’s audited accounts?

    Open
  3. 03

    How much arose from premiums, assay charges, service fees, off-taker discounts, trading losses and exchange-rate movements?

    Open
  4. 04

    What additional financing and sterilisation costs arose from the programme?

    Open
  5. 05

    What portion of the costs arose before and after GoldBod became operational?

    Open
  6. 06

    What proportion of reserve accumulation can reasonably be attributed to the programme, and what would comparable accumulation have cost through conventional operations?

    Open
  7. 07

    What borrowing costs, collateral arrangements, foreign-exchange exposures or government guarantees accompany the new financing model?

    Open
  8. 08

    Will Parliament, the Auditor-General or another independent institution undertake a consolidated reconciliation of Bank of Ghana, GoldBod and government accounts?

    Open

GPEx policy judgement

Red/Amber: substantial verified risk; corrective restructuring underway.

Ghana’s Domestic Gold Purchase Programme generated very substantial financial costs. The evidence does not establish that GoldBod itself made a US$1.7 billion corporate loss. Programme benefits and significant costs can coexist, and the consolidated net economic impact remains unverified. GPEx recommends an independent reconciliation of Bank of Ghana, GoldBod and relevant government accounts for Parliamentary and public scrutiny.

Substantial verified monetary and fiscal-governance risk · corrective restructuring underway · consolidated net economic impact not yet verified

Source trail

Primary records and operating-model evidence

Sources establish the reported amounts, institutional accounts and governance response; they do not yet provide the required consolidated reconciliation.