Our use of digital tools for simplification is consistent with our recent mandate to spur the creation of “critical policy audiences” within the broad Ghanaian public to stem the “katanomic crisis” in Ghanaian policymaking.
The anti-katanomic modelling dashboard above should help the public appreciate that while the GoldBod-G4R program has successfully centralized gold flows, and pushed the bulk of gold dollars away from other players in the economy (such as Banks), it has done so by creating a pro-cyclical “bear trap.”
Meaning: the system is heavily leveraged on historically high gold prices. Should prices correct, the cost of maintaining the same level of FX intervention would require even greater liquidity injections and higher sterilization costs, precisely when the asset backing the operation is losing value.
The GoldBod-benefits paper is being positioned as an expert-academic vindication of the program. However, a granular scrutiny of the methodological choices reveals significant biases that render the paper a policy defense document rather than an independent audit.
Some of this bias is discernible in the heavy reliance on the “Formalisation Shock” hypothesis. The authors observe that recorded ASM gold exports surged from 63.6 tons in 2024 to 103.0 tons in 2025 and attribute this entire increment (39.4 tons) to the successful formalization of previously smuggled domestic gold. They value this increment at $3.8 billion and count it as a net benefit.
Appiah has already shown by simple reference to the historical trend of ASM gold production in Ghana that we have seen massive jumps in output in the past that even exceeds the 2024 – 2025 leap. But the attribution suffers from another big gap: it ignores the regional arbitrage factor. The paper fails to control for the possibility that the “GoldBod Bonus” (pricing gold above world market rates) has incentivized the smuggling of gold into Ghana from neighboring Burkina Faso, Mali, and Côte d’Ivoire. This is a phenomenon corroborated by independent analysts.
If Ghana pays a premium, it becomes a magnet for regional gold. Counting this “transit gold” as a triumph of domestic formalization is a categorical error. It flatters the export data but hides the fact that the BoG is effectively subsidising regional miners with Ghanaian public funds.
We will return to these points in detail again as they are so important.
A Benefit-Cost Fallacy: Gross Inflows versus Net Economic Cost
The paper’s central quantitative argument is a cost-benefit analysis where “benefit” is defined as the gross value of FX inflows ($3.8 billion) and “cost” is limited to the reported trading loss ($214 million), yielding a ratio of ~18:1. This is mind-bogglingly naïve.
By means of such economic sleight-of-hand, the authors gloss over the fact that Gross FX inflow is a liquidity metric rather than a welfare metric. The true economic benefit is the value added by that liquidity, minus the full cost of acquiring it. By excluding the massive costs of sterilization (interest on BoG bills) and the opportunity cost of the capital employed, the authors present a grossly distorted picture. They essentially argue that buying $1 for $1.05 is a success because you now have $1 in hand. This logic would bankrupt any private entity; applying it to sovereign resources is nothing short of dangerous.
The narrative of GoldBod is that it “captures” gold that was previously lost. However, the data shows that while volume passing through the BoG has increased, the cost of capturing that volume has skyrocketed.
In 2021, the BoG received gold dollars largely as a regulatory requirement (surrender portion) from large mines at zero marginal cost to the state (other than the opportunity cost of allowing retention). In 2024/2025, the BoG is buying the gold. It is swapping Cedi assets for Dollar assets.
The shift from a regulatory surrender regime to a direct purchase regime means the state has moved from a passive beneficiary to an active, risk-taking trader. The “increase” in inflows is funded by debt (BoG bills), rather than by organic economic growth.
The mistake that most members of the public make is to assume that the GoldBod is somehow responsible for the creation of $10.3 billion in new value for the economy. They forget that the government is simply aggregating Dollars from miners and traders and then selling it back to importers, sometimes at a subsidised rate due to the aggressive peg of the Cedi to the Dollar. There is no “new value” being created per se.
The easiest way to appreciate this point is to ignore, for one moment, the Bank of Ghana’s reserves story and focus on what has been happening at the commercial banks.
Between 2019 and 2024, forex transactions at the commercial bank window increased by more than 400%. Entirely market-driven. In the same vein, transactions more than halved between 2017 and 2019. In a system capable of such massive forex swings. It takes a lot of chutzpah to reduce the forex supply channel of impact on the exchange rate to just the supply of gold-dollars from GoldBod for episodic BoG interventions. A much more elaborate analysis is required before the authors of the GoldBod-benefits can make the claims they did in their paper.
Of course, during a crisis period where disorderly conditions in the market needs to be tamed, the central bank’s “smoothing” power requires a concentrated forex stash. What we must not do is confuse a tool for emergency stabilisation with an optimal routine policy.
Historically, UAE import data for gold from Ghana has consistently exceeded Ghana’s reported exports by billions of dollars, providing analysts with the best empirical signal for “smuggling”.
Unfortunately, some have elevated this useful gap-filler into an all-round justification for GoldBod’s anti-smuggling prowess even though the 2025 UAE-Ghana discrepancy data for Ghana is not officially out (fragments of that data exposed in confidence to this author for certain periods in 2025 does suggest that such gaps may still exist by the way).
Yet, the presence of such gaps can result from a confusing set of factors such as import/export valuation, period-bounding, routing/origin classification issues, product/HS-scope mismatches, timing mismatches, and/or reporting/valuation differences tied to contract structure and customs valuation.
Below, to illustrate the above point, I present data I pulled from the data portal of the Indian Ministry of Commerce & Industry for the gold trade between Ghana and India.
At first glance, it shows persistent gaps between import and export data covering the gold trade between the two countries. It would be naïve to describe this as evidence of smuggling without painstaking granular analysis.
The Swissaid report that everyone relies on when analysing the Ghana – UAE discrepancies note also that the discrepancies on the Ghana – India leg are very minimal (see historical data below). And, yet, as clearly demonstrated above, those discrepancies appear to have taken a wild turn since the Domestic Gold Purchasing Program intensified in late 2024.
Allow me to present a range of deeper considerations one need to keep in mind when attempting to analyse discrepancies of the type the GoldBod-benefits paper strives to latch the smuggling thesis on.
Without anything approaching that level of granularity and sophistication in analysis, the GoldBod-benefits paper proclaims the GoldBod as having extinguished smuggling networks and formalised previous smuggling networks.
Loss Dynamics
As we have already shown elsewhere, the trading losses recorded by the GoldBod aren’t due merely to “translational accounting effects” as the GoldBod-benefits paper’s authors sought to suggest. The losses are structural, and embedded in the very design of the G4R pricing model.
To displace the efficient private network of buyers, GoldBod must offer a price that clears a distorted market, inclusive of a “bonus.”
Combining the buying premium and selling discount for trades of that nature reveals the structural rot.
For every $1 billion of gold traded, the state can easily lose between $53 million and $97 million in direct trading value. This explains the IMF’s $214 million figure. It is not a “paper loss”; it is the cost of buying market share.
Yet, the GoldBod-benefits paper twists its knickers into knots in an attempt to underplay such a fundamental issue.
The BoG does not buy gold with tax revenue. It buys gold by either, in a minority of situations, managing swaps with commercial banks in which the banks have no downside risk because the central bank absorbs all of it, or, in the majority of cases, crediting the accounts of aggregators with newly created Cedis. For a batch of 250,000 ounces at ~$4,255/ounce (approximately GH¢12.1 billion), the monetary base (M0) expands instantly.
In an inflation-targeting regime, albeit one distorted by an undisclosed secondary anchor (the currency peg), such a massive injection of high-powered money is dangerous. To prevent inflation from spiralling (which would weaken the cedi the program is trying to save), the BoG must “sterilize” this injection. It essentially takes the money back out of the system by selling BoG Bills to banks.
The cost of this operation is the interest paid on the bills, as mentioned several times in the past.
These OMO costs are quasi-fiscal. They function like government borrowing but do not appear in the budget deficit reported by the Ministry of Finance. They appear as “operating losses” on the BoG’s balance sheet and will likely exacerbate the negative equity position of GH¢61.32 billion reported at the end of 2024. This is a hidden debt accumulation mechanism that future taxpayers will eventually have to recapitalize.
I find it very hard to fathom why the GoldBod-benefits authors would completely ignore all these entries in the ledger and yet purport to be engaged in a dispassionate analysis of costs and benefits.
Of course, no mention is made of the more than $2.5 billion the IMF has injected into the economy when the issue of inflows comes up.
But surely, even a paper focused on making GoldBod look good can still try to engage more analytically with a few of the other drivers of the BoG’s reserves accumulation trajectory?
In January 2024 Ghana agreed an Agreement-in-Principle (AIP) with its Official Creditor Committee (Paris Club countries and China, among a few others) covering about US$5.17 billion of bilateral loans (cut‐off Dec 2022). Under the agreed terms, Ghana will receive roughly US$2.85 billion in debt‐service relief over the 2023 – 26 time period. This reflects lower principal repayments and interest (e.g. through extended maturities and grace periods).
Separately, Ghana finalized a debt exchange with private Eurobond holders. In October 2024, it restructured US$13.1 billion of sovereign bonds. Creditors took a ~37% nominal haircut, reducing the outstanding principal by about US$5 billion and cutting the average coupon from ~8% to ~5%. The Eurobond deal yields roughly US$4.3 billion of debt‐service savings during the 2023 to 26 timeframe (both principal and interest).
In total, therefore, Ghana’s external debt restructurings are projected to cut external debt‐service obligations by about US$9.68 billion over the medium term (2023 to 26), which is roughly US$2.7 to 2.9 billion per year on average.
What is important is that those savings compound as each new agreement is reached and terms become effective. Those amount to forex obligations that the Bank of Ghana no longer needs to release dollars to support the government in payouts to creditors.
At the same time that the Bank of Ghana’s dollar obligations have diminished, the country is facing a significant slowdown in secular imports growth that has led to a trade surplus of roughly $8.5 billion in recent times. Essentially, Ghana’s goods exports generated $8.5 billion more than the country needs for goods imports. Of course, one must account also for the service imports and other payment obligations, but a good look at the current account position shows that all these have moderated in recent times.
Then there is the general effects of sterilisation (beyond the cost issues we have raised), which removes Cedis that would have otherwise gone chasing Dollars from circulation.
The GoldBod benefits paper casually browses through such possible contributory elements but it makes virtually no effort to quantify their differential contribution to the relative stability.
Because the sterilisation and liquidity management issues are so critical in appreciating the full life-cycle of the gold-for-reserves policy and the GoldBod mechanism that underwrites it, we have dedicated the brief section that follows exclusively to explaining our stance.
The Sterilisation Bogey
Under a standard Inflation Targeting (IT) framework, which is the dominant posture of monetary policy in Ghana, the Central Bank sets a Monetary Policy Rate (MPR) to anchor inflation expectations. It manages the interbank interest rate to align with this MPR using Open Market Operations (OMO). Liquidity in the banking system is influenced by two primary autonomous factors:
The sum of NFA and NDA constitutes Reserve Money (RM) or High-Powered Money.
ΔRM = ΔNFA + ΔNDA
If the increase in RM exceeds the level consistent with the inflation target, the Central Bank must “sterilize” the excess. It does this by selling BoG Bills to commercial banks. This operation reduces the banks’ excess reserves but creates an interest-bearing liability for the Central Bank. The interest paid on these bills constitutes the Sterilization Cost or OMO Cost.
The mechanics of the Gold-for-Reserves program related to this issue are as follows:
To discern the effect of this process on the reserve money trajectory in Ghana requires, first of all, an accounting for baseline effects. Because the response to the COVID-19 pandemic involved a massive surge in Cedi liquidity, a naïve tracking of reserve money would fail to capture the scale of sterilisation associated with the domestic gold purchasing programs.
Moreover, in 2023, Ghana entered into a US$ 3 billion Extended Credit Facility (ECF) with the IMF. A core conditionality was the cessation of monetary financing of the fiscal deficit. This shifted the driver of liquidity from fiscal injection (NDA) to reserve accumulation (NFA) and required the Bank to aggressively mop up the existing liquidity overhang.
Thus, the Bank of Ghana’s 2024 financial statements reveal that the cost of Open Market Operations reached GH¢ 8.60 billion. A secondary cost of operation, currency printing, rose to GH¢ 1.01 billion in 2024 from GH¢ 0.69 billion in 2023, reflecting the nominal expansion of money demand. The end-result was that Total Operating Income (GH¢ 9.40 billion) was completely consumed by Total Operating Expenses (GH¢ 18.89 billion). The GH¢ 8.60 billion OMO cost accounted for 45.5% of total operating expenses, confirming it as the single largest driver of the central bank’s deficit.
In 2024, the Bank absorbed a total of GH¢ 134 billion in liquidity through its operations.This massive figure underscores the scale of the sterilization required to counteract the liquidity injected by gold purchases and other autonomous factors.
Despite this massive absorption, reserve money continued to grow, driven entirely by the NFA component (gold and FX accumulation). The Bank successfully kept reserve money within the IMF program’s indicative targets, preventing a resurgence of hyperinflation.
The only thing that has abated the costs somewhat in 2025 is that the policy rate fell throughout the year. The reduction in the rate implies a significant deceleration in the accrual of OMO costs in Q4 2025. Based on the run-rate of GH¢ 8.6 billion in 2024 and the significant rate cuts in late 2025, the total OMO cost for 2025 is projected to moderate to the range of GH¢ 6.5 – 7.5 billion. This should not however deceive observers into underestimating the scale of sterilisation operations involved. As of July 2025, according to briefings by the Governor of the Bank of Ghana to Parliament, the volume of sterilisation already exceeded GH¢ 60 billion.
Triangulating the policy rate reduction, the OMO costs projected for 2025, and the reserve money growth rate consistent with the IMF program, yields a composite view of an aggressive level of sterilisation operations, perhaps unprecedented in recent times.
The paper says nothing about comparative experiences elsewhere
Comparing Ghana’s GoldBod to peer programs highlights important divergences in design and outcomes:
All three programs (Ghana, Philippines, and Mongolia) share the goal of converting domestic gold into reserves. But Ghana diverges in execution by introducing strange variables.
As outlined in the table below, Ghana’s approach exposes serious tensions with best practices such as monopoly versus market, proximity to central bank funding versus arm’s-length financing, and low transparency versus international standards.
These differences have real consequences: Ghana’s gold flows are less diversified ($10 billion in ~11 months largely to a few buyers in India and the UAE), whereas the Philippines and Mongolia maintain multiple international off-takers to mitigate counterparty risk.
Ghana also seems to have pegged its demand for central bank gold dollars at an unrealistically high level due to a desperate need to over-intervene in the FX market.
We have proposed the Trust-Chain model to preserve the benefits of Ghana’s gold program, which include contributing to reserves accumulation that supports currency stability and formalization of small-scale mining. While at the same time eliminating the inefficiencies stemming from state-dominated trading. This is what we consider to be crucial.
By moving GoldBod “upstream” into a governance and oversight role, and empowering the market to handle trading under transparent rules, the government can still achieve its objectives without directly footing the entire bill. The Trust-Chain would do exactly that: make currency stabilization via gold a shared responsibility of the central bank, finance ministry, private investors, and the mining community – rather than a costly solo venture.
It goes without saying that such a model would only be viable if the Bank of Ghana is not desperately suctioning all the gold in Ghana in a relentless intervention spree to fix the currency at an unrealistic level. Smart modelling is required to better and transparently anchor the right currency fluctuation band (responsive to market conditions), the rational degree of intervention, and the consequentially appropriate volume of gold-dollars the supply of which the central bank must backstop to meet policy goals.
Through the simulations we have started to roll out, it becomes clear that there’s a policy tightrope: Ghana can push certain levers to gather more gold and seek to stabilize the currency, but pushing those same levers too far inflates the cost of the intervention. The dashboard-widget we have implemented visually underscores the core trade-off: aggressive currency support vs. financial losses.
The overriding analytical imperative that analysts and scholars truly committed to a win for Ghana should be focusing on is a program to model the reasonable band of cedi to USD value that warrants intervention, the volumes of gold realistically needed to maintain fluctuation within that band through intervention, and, consequently, the acceptable level of losses justified by the policy benefit, bearing in mind that the performance of the Cedi is not dictated solely by reserve position. Furthermore, such analysis also exposes the scenarios under which gold-dollar driven intervention and its associated costs could become a net negative.
In charting a course to refine its Gold-for-Reserves program, Ghana has the opportunity to set a powerful example. It can demonstrate how a developing country, through ingenuity and reform, turned a quick fix into a permanent feature of economic resilience. Done right, the model could become a blueprint for others. The path could be lighted as to how to leverage natural resources for national stability without succumbing to the pitfalls of opacity or unsustainable costs.
With prudent reforms and an unwavering commitment to transparency, Ghana can strike gold in more ways than one. By securing more than just reserves, the country’s leaders can build trust and wealth for generations to come. Those of us in the knowledge and civil dialogue space – as analysts, academics, and commentators – should be focused on driving higher performance in this fashion instead of generating feel-good vibes for the government through incomplete analysis.
