What Sound Money for Ethiopia Would Actually Require

The prescription

A series that has spent twenty-six articles diagnosing a disease owes its readers, at the end, a prescription. This article provides one. It is not a programme; it is a set of institutional requirements that, taken together, would constitute what the series has been calling sound money for Ethiopia. The requirements are not original to this article — most of them appear, in some form, in the IMF Article IV consultations of the past decade, in the academic literature on central-bank reform, and in the comparative cases article 24 examined. The article’s contribution is to assemble them into a coherent package and to be specific about what each would mean in Ethiopian conditions.

The package has six components. None is sufficient on its own. Together, they constitute the architecture that would make the disease this series has been tracking structurally harder to recur. Without them, the post-2024 reforms are reversible. With them, they are durable.

The honest caveat: implementing all six is politically difficult, would impose substantial transition costs, and would require sustained political commitment over more than one electoral cycle. The article is not arguing that implementation is easy; it is arguing that implementation is necessary, and that the difficulty of implementation is the reason the disease has been as durable as it has. There are no shortcuts.

Component one: a genuinely independent central bank

The first and most important component is a National Bank of Ethiopia with operational autonomy on monetary policy, statutory protection from executive direction, and binding limits on lending to the government that cannot be lifted by simple political decision.

The 2024–25 reforms of the NBE statute, examined in articles 21 and 22, represent a partial step toward this. The new law strengthens the bank’s formal independence and prohibits monetary financing of government deficits. The IMF’s own assessment is that the law is “a significant advance” with “remaining gaps with respect to governance and autonomy.”¹ The remaining gaps are precisely what would have to be closed.

Specifically: the appointment process for the NBE governor and senior leadership must be insulated from executive discretion. The governor should serve a fixed term that does not coincide with the political cycle; removal should require a specified procedure rather than executive prerogative. The NBE Board should include non-government members with explicit fiduciary obligations to the bank’s mandate rather than to the government. The bank’s budget should be controlled by the bank itself rather than by the Ministry of Finance. The bank’s policy decisions should be made by a committee operating under transparent procedures, with published meeting minutes that allow public scrutiny.

The Rwandan central bank, as article 24 examined, has substantial portions of this architecture. So do most of the central banks of the developed world. The Ethiopian gap is, in the technical sense, knowable and addressable.

What would not work — and what previous reforms have, in various forms, attempted — is a formally-independent central bank that the executive can override in practice through informal political pressure on a governor who serves at the executive’s pleasure. The formal architecture has to be backed by the institutional infrastructure that makes the formal protection operational.

Component two: a binding fiscal rule

The second component is a fiscal rule that limits the federal government’s deficit to a level the country can finance through means other than central-bank lending. The rule has to be: stated in clear terms, embedded in law or constitutional provision sufficient to make change difficult, monitored by an independent fiscal council with publication authority, and enforced through political-cost consequences for breaches.

A reasonable Ethiopian fiscal rule would target a federal deficit not exceeding 3 percent of GDP in normal years, with escape clauses for specific emergencies (natural disasters, major external shocks) subject to subsequent restoration. The deficit would have to be financed through some combination of tax revenue, domestic non-bank borrowing (through the Ethiopian Securities Exchange), and concessional external borrowing. Central-bank lending to the federal government would be prohibited as a financing source.

The rule would, by construction, force the government to either expand the tax base or compress expenditure when fiscal pressures arose. This is not a comfortable choice. It is the choice that disciplined fiscal management requires. The Chilean fiscal-responsibility framework, the German constitutional debt brake, the Brazilian fiscal-responsibility law of 2000 — each provides a model for how this can be constructed institutionally in countries that have grappled with similar problems.²

The Ethiopian political-economy challenge, the series has argued throughout, is that the fiscal-monetary linkage has been used to fund patronage flows that the political coalition has not been willing to give up. A binding fiscal rule would close this option, forcing the coalition to fund patronage through visible mechanisms or to give it up. The cost of imposing the rule will be felt principally in the political space that the patronage flows have occupied. The benefit will be felt by the population whose welfare the monetary disorder has been eroding.

Component three: current-account convertibility

The third component is full convertibility of the birr for current-account transactions, with a market-determined exchange rate, and the elimination of the forex rationing apparatus that has characterised Ethiopian monetary policy for fifty years.

The 2024 float was the foundation of this component. The next steps are the elimination of remaining current-account restrictions: surrender requirements on exporters (substantially reduced in 2024–25 but not yet fully eliminated), import restrictions and priority lists, the residual capital controls on trade-related transactions. The post-float regulatory framework, as it has evolved through 2025–26, is moving in this direction but is not yet complete.

The capital-account side is more contested. Standard development-economics arguments support gradualism on capital-account liberalisation — full convertibility for current account first, capital-account liberalisation sequenced over a longer period as the institutional infrastructure develops. The series accepts this argument. The current-account convertibility is the priority; capital-account liberalisation can be sequenced.

What is not negotiable, in the series’s framing, is the market-determined exchange rate. The post-2024 regime has, in practice, moved closer to this; the next test will be whether the NBE accepts continued exchange-rate adjustment when it is politically uncomfortable. The temptation to use NBE foreign-currency auctions to target a specific exchange-rate level rather than to manage liquidity will be substantial, particularly during stress periods. The discipline required is to resist the temptation. The institutional architecture of component one is what would make that resistance possible.

Component four: deep capital markets

The fourth component is the development of domestic capital markets capable of financing Ethiopian public and private investment in birr, rather than requiring continuous foreign-currency borrowing.

The Ethiopian Securities Exchange launched in 2024 is the institutional foundation. Building the market into a substantial source of financing will take years. The required components include: a sovereign-bond market with regular issuance schedules, deep secondary trading, and price-discovery functions that allow the term structure to be observed and used; a corporate-bond market that allows large private and parastatal issuers to fund investment through market-priced debt; an equity market with sufficient breadth that institutional investors can construct diversified portfolios; the regulatory infrastructure (Capital Market Authority, custodial arrangements, settlement systems) that supports the markets and protects investors.

The development of these markets requires demand and supply sides to coalesce. The supply side requires issuers willing to come to market at the rates the market will offer; this requires macroeconomic stability that gives issuers confidence in birr-denominated obligations. The demand side requires institutional investors (pension funds, insurance companies, banks beyond their normal commercial activity) with the regulatory mandate and the operational capacity to participate in the market. Both sides are at early stages.

The relevance to sound money is that a country with deep birr-denominated capital markets has alternatives to central-bank financing of fiscal pressures. The fiscal rule of component two is enforceable only if there is somewhere other than the central bank where the government can fund its needs. The capital market is that somewhere. Without it, the fiscal rule is, in stress periods, an invitation to seek the central-bank alternative.

Component five: real land collateral

The fifth component is the reform of the land regime to allow Ethiopian land to serve as collateral for credit, in the way article 13 examined.

The constitutional commitment to state ownership of land is, the series acknowledges, embedded in Ethiopian political consensus in ways that make outright privatisation politically infeasible. What is feasible is the development of long-term transferable leasehold — a regime in which land remains state-owned but use-rights are formally registered, long-term, transferable, and pledgeable as collateral. The Chinese system since 1988 operates on this basis and has produced both functional property-rights infrastructure for credit purposes and continued state ownership.

The Ethiopian Rural Land Certification programme, examined in article 13, has issued certificates to millions of farmers but has not made the certificates pledgeable as collateral. The reform that would make them pledgeable is institutionally manageable: changes to banking regulation, changes to land-administration regulations, the development of court infrastructure for foreclosure proceedings. The political resistance is the binding constraint.

The implication for sound money is direct. Component four (deep capital markets) is constrained by the absence of land collateral, because the private credit market that would feed capital-market activity cannot expand without something to lend against. The chain runs from land collateral through private credit through domestic capital markets through fiscal discipline through monetary stability. Breaking the chain at any point — and the land link is the most binding — makes the rest of the package harder to implement.

Component six: a depoliticised banking system

The sixth component is the depoliticisation of the Ethiopian banking system: privatisation of the Commercial Bank of Ethiopia (in full or in significant part), restructuring of the Development Bank of Ethiopia along commercial lines, the elimination of directed-lending requirements on private banks, and the entry of foreign banks operating under standard regulatory frameworks.

The 2024–25 reforms have begun this process. Foreign-bank licensing is being put in place. The CBE’s commercialisation has been announced. The 27 percent rule is gone. But the deep depoliticisation has not happened. The CBE remains the dominant institution. State-favoured borrowers continue to receive preferential treatment. The political pressure on lending decisions continues, even where its specific instruments have been removed.

What would complete the depoliticisation: substantial private-investor participation in CBE through either an IPO or strategic-investor sale, with the government retaining a minority stake or none at all; restructuring of DBE into either a fully-commercial entity or a narrowly-mandated development-finance institution with explicit subsidies that appear on the formal budget rather than being hidden in below-market interest rates; entry of multiple foreign banks with substantial market share, providing competitive pressure on the legacy institutions; regulatory enforcement that operates on the same standard for all participants regardless of ownership.

The connection to sound money is the connection of components four through six together. A banking system that operates on commercial principles, lending against real collateral, in the context of deep capital markets and independent monetary policy, would not require the financial-repression machinery the series has been tracking. The machinery would be unnecessary because the function it has performed — funding fiscal pressures and state priorities — would be performed by the visible mechanisms of taxation and bond issuance. The disease’s mechanism would be absent.

The political precondition

These six components do not implement themselves. They require a political coalition that is committed to them as core policy objectives — not merely as conditions imposed by external creditors, but as the coalition’s own programme.

The series’s argument throughout has been that no such coalition has existed in Ethiopia across the period covered. The post-2024 reforms have been adopted under external pressure (IMF programme, sovereign-debt crisis, the need for World Bank financing) and the political coalition has accepted them more as a constraint than as a commitment. The reforms are, on this analysis, vulnerable to reversal when the external pressure relaxes.

What would change this? The series’s editorial position is that the political coalition for sound money will not emerge through technocratic articulation alone. It will emerge through political mobilisation by the constituencies who pay the cost of the existing arrangement: the urban worker whose real wage has been destroyed, the smallholder farmer whose remittance income loses value year by year, the pensioner whose retirement provision has been eroded, the small business owner who cannot access credit. These constituencies have, historically, lacked the political organisation to demand reform. The post-2022 inflation episode has, the series predicts, created the conditions for their political mobilisation. Whether the mobilisation occurs, and whether it produces a coalition committed to the six components above, is the open question of the next decade.

The next and final article synthesises the series’s argument and turns to that question of political mobilisation directly.

---

For central-bank-independence frameworks, see Alex Cukierman, Central Bank Strategy, Credibility, and Independence (MIT Press, 1992); Lars E.O. Svensson, “Inflation Targeting,” in Handbook of Monetary Economics, Vol. 3B (2011).

For the Chinese leasehold land system, see Klaus Deininger and Songqing Jin, “The Potential of Land Markets for Productivity, Welfare, and Equity: Evidence from China,” World Bank Policy Research Working Paper 5345 (2010).

For the literature on capital-market development in low-income countries, see Aslı Demirgüç-Kunt and Ross Levine, eds., Financial Structure and Economic Growth (MIT Press, 2001); Asli Demirgüç-Kunt, Erik Feyen, and Ross Levine, “The Evolving Importance of Banks and Securities Markets,” World Bank Economic Review 27, no. 3 (2013).

Footnotes

1. International Monetary Fund, “Ethiopia: Second Review under the Extended Credit Facility Arrangement,” IMF Press Release 25/006, January 2025. ↩

2. For the comparative literature on fiscal rules, see Xavier Debrun, Manmohan S. Kumar, et al., Fiscal Rules in Response to the Crisis, IMF Working Paper, 2008; George Kopits, ed., Rules-Based Fiscal Policy in Emerging Markets (Palgrave Macmillan, 2004). ↩