In 2025, the cedi posted its first annual gain against the US dollar in more than three decades, while the Bank of Ghana posted a combined loss of GHS 34.95 billion. Read against the Plaza and Louvre Accords, the two outcomes are not contradictory: they are the price of stabilising a currency alone.

Introduction

In 2025, the Bank of Ghana (BoG) accomplished something of extraordinary monetary significance. The cedi appreciated by 40.7 per cent against the US dollar, its first annual gain in more than three decades and, on IMF data, the strongest performance of any African currency. Headline inflation, which had peaked at 54.1 per cent in late 2022, fell to 5.4 per cent by year-end, and gross international reserves rose to a record US$13.83 billion.

The same year produced large losses on the Bank’s balance sheet. Its audited statements, released on 30 April 2026, show a net operating loss of GHS 15.63 billion, up from GHS 9.49 billion in 2024, and a further loss of GHS 19.32 billion in other comprehensive income. The combined loss of GHS 34.95 billion pushed cumulative negative equity to about GHS 96.3 billion.

Severed from their context, these figures read as catastrophe. Situated within the history of coordinated exchange-rate statecraft, they read as something else entirely: the quantified cost of macroeconomic stewardship in the absence of multilateral burden-sharing.

The modern international monetary system rests on a structural paradox. Capital moves across borders in quantities that dwarf the reserves of any single monetary authority, yet central-bank mandates remain resolutely national, so that misalignments are increasingly absorbed on central-bank balance sheets. At decisive moments, the major economies answered such misalignments with coordinated intervention, whose efficacy is usually traced to the signalling channel.

A unilateral stabiliser lacks that multilateral credibility by definition. Yet Ghana’s outcomes suggest that unilateral intervention can replicate much of what coordination achieved, provided it is sufficiently aggressive, sufficiently sustained and anchored by credible external surveillance. The price is borne in full by one institution and recorded with brutal clarity on its balance sheet.

The lesson of Plaza and Louvre

The Plaza Accord of 22 September 1985 remains the defining act of post-Bretton Woods monetary diplomacy. By the spring of that year, the dollar had appreciated by roughly 44 percent against a trade-weighted basket over five years. Its causes were well understood: American fiscal expansion, colliding with a tight Federal Reserve, had drawn in capital through real interest-rate differentials, while a widening trade deficit fed protectionist pressure in Congress.

The finance ministers and central-bank governors of the G-5 (the United States, Japan, West Germany, France and the United Kingdom) responded with a communiqué of studied ambiguity but unmistakable intent. The dollar was overvalued, and they would cooperate to encourage its orderly decline through concerted intervention.

Their dollar sales, about US$10 billion in the following weeks, were modest by modern standards. Their power lay in the signal. Sterilised intervention cannot durably move exchange rates through portfolio-balance effects alone; its force derives from the credible public commitment of several sovereigns, each bearing reputational and financial cost. That jointness, constitutive rather than merely logistical, is precisely what a unilateral stabiliser cannot reproduce.

In the event, Plaza overachieved: by early 1987, the trade-weighted dollar had fallen by about 40 percent from its peak, and an excessively weak dollar had displaced the original problem. The Louvre Accord of February 1987 reversed the objective, seeking to halt the slide within informal reference ranges.

The episode showed that defending an equilibrium is structurally different from engineering a correction. Defence requires continuous sterilisation, sustained liquidity management and interest-rate differentials wide enough to deter speculation, and each carries a direct cost to the central-bank balance sheet. The participating central banks absorbed those costs collectively.

Neither communiqué said who would pay. Exchange-rate operations are quasi-fiscal in character, redistributing income and imposing costs much as explicit taxation or public expenditure would. IMF staff research notes that such losses are met from equity buffers where these suffice, or otherwise through government transfers; either way, they are fiscal costs. The G-5, with deep fiscal capacity and implicit mutual backstops, socialised those costs across their treasuries. Ghana had no equivalent mechanism.

The solitary stabiliser

Between 2022 and 2025, the BoG performed the correction-and-defence function of the Plaza–Louvre model entirely alone. It did so under a US$3 billion IMF Extended Credit Facility, in the aftermath of the most severe domestic debt restructuring in Ghana’s post-independence history, and against a legacy of fiscal dominance and structural excess liquidity.

When advanced economies tighten aggressively, capital flight and dollar strength exert depreciatory pressure on African currencies, feeding imported inflation and swelling the cost of foreign-currency debt. The cedi lacks the reserve status that the yen and the deutschmark enjoyed in the 1980s, yet the Bank of Ghana Act, 2002 (Act 612) mandates price stability and support for a stable path of the cedi. The Bank must therefore intervene alone, trading institutional equity for currency stability.

The research from which this article is drawn names such an institution a solitary stabiliser, defined by three features:

  • It intervenes alone, so the entire cost of stabilisation accrues to one balance sheet.
  • It lacks a multilateral signal, so its credibility must be manufactured through the persistence of its own commitment.
  • It has no formal guarantee of recapitalisation, so its losses accumulate against an implicit and untested promise of fiscal support.

The Mundell–Fleming trilemma holds that no country can combine a fixed exchange rate, free capital mobility and monetary independence, and Hélène Rey argues that the global financial cycle reduces it to a dilemma. Neither fully captures the BoG’s position. The Bank faced a quadrilemma, in which four objectives could not all be maintained together:

  • exchange-rate stabilisation;
  • tight monetary policy for inflation control;
  • fiscal consolidation under the IMF’s zero-financing rule; and
  • central-bank financial health.

Against the structural excess liquidity associated with the Domestic Debt Exchange Programme (DDEP), large policy-driven losses were not merely probable but arithmetically inevitable. Ghana was not choosing between the corners of the impossible trinity; it was attempting to occupy three adjacent policy positions at once. The losses were the clearing mechanism through which that overdetermined system resolved its contradictions, with the balance sheet as absorber of last resort.

Credibility came, paradoxically, from an external constraint. The sequence of IMF programme reviews, each demanding performance on inflation, reserves and fiscal targets, formed a serial credibility mechanism that the BoG could not have generated on its own. For central banks without the geopolitical capital of G-5 membership, IMF conditionality, deployed as a substitute commitment device, is thus the closest functional equivalent of multilateral coordination. It carries sovereignty costs, but it can deliver outcomes comparable in quality, if not in mechanism, to coordinated intervention.

Three sources of loss

The headline loss is not a homogeneous expression of failure. It arose from three structurally distinct mechanisms.

The first and largest is the cost of open market operations (OMO), which reached GHS 16.73 billion in 2025, nearly double the GHS 8.60 billion of 2024. It is best understood as a monetary tribute: the price a post-crisis financial system, already distorted by the DDEP, exacts for withdrawing the excess liquidity that crisis-era monetary financing created.

Launched in December 2022, the DDEP required domestic bondholders to swap existing securities for longer-dated, lower-yield bonds. It imposed impairments of about GHS 48 billion on the BoG’s holdings, including a 50 percent haircut on non-marketable instruments, and cut its annual interest income by about GHS 13 billion until the affected instruments mature.

The restructuring also steered commercial-bank liquidity into BoG money-market instruments, which offered attractive short-term returns amid still-elevated policy rates. By 2025, the OMO book stood at about GHS 93.6 billion, at rates between 18 and 28 percent. Without sterilisation, that liquidity would have fed inflation and depreciation and unwound gains already won. The OMO bill was the price of not losing them.

The second, GHS 9.05 billion in losses on gold transactions, is the most mischaracterised component. Under its domestic gold-purchase programme, run alongside the Ghana Gold Board (GoldBod), the Bank bought locally mined gold from artisanal and small-scale miners in cedis and converted it into reserves. The loss arises because the cedi acquisition cost, recognised at purchase, exceeds the carrying value at which the gold enters the balance sheet. It is a timing mismatch, not a destruction of value.

At the end of 2025 the Bank still held about 18.6 tonnes of gold, worth roughly three times the loss attributed to the programme, having sold some 19 tonnes at a realised profit.

The strategic logic reaches beyond reserve accumulation. By requiring domestic gold to be settled in cedis before export, Ghana’s gold-governance architecture routed foreign-exchange earnings through official channels, reduced illicit financial flows and gave the BoG a continuous source of dollar liquidity for intervention. The programme was instrumental in lifting gross international reserves from US$9.11 billion to US$13.83 billion.

The third is an appreciation paradox, in which Ghana’s most celebrated achievement generated large losses. When a central bank holds foreign-currency reserves and its own currency appreciates, the local-currency value of those reserves falls mechanically. A US$1 billion holding is worth GHS 17 billion at 17 cedis to the dollar, but only GHS 10 billion at 10.

In 2025, the cedi’s rise produced GHS 5.47 billion in revaluation losses in the operating account and GHS 19.32 billion in other comprehensive income. This is a structural feature of central-bank balance sheets in small open economies, not a policy error. The Bundesbank and the Bank of Japan experienced exactly this effect after Plaza, as their dollar sales lifted the deutschmark and the yen. The economic substance is structurally equivalent; only the distribution of the burden differs.

The effect is also reversible, as Governor Johnson Pandit Asiama has noted. By May 2026, the cedi had weakened to about GHS 11.5 to the dollar from the GHS 10.4 at which the 2025 accounts were struck, the signature of a translation effect rather than a real depletion of reserves.

Ghana among the loss-making stabilisers

Ghana’s experience is far from unique, and the Bank for International Settlements (BIS) has been emphatic that losses and negative equity do not in themselves impair a central bank’s capacity to operate. The European Central Bank (ECB) lost €7.94 billion in 2024 by paying elevated interest on banks’ reserves while earning low yields on legacy bonds, a structure that closely mirrors the BoG’s OMO costs. The ECB carries these losses forward against future profits, with no recapitalisation requirement.