Africa’s total public debt has risen more than fourfold since the early 2000s, but just as important as the increase in debt volumes is the shift in structure. This column uses a new open-access dataset covering more than 50,000 loans and securities issued by 54 African countries to reveal that African governments now raise more than half of their financing at home, reversing decades of dependence on external lenders. While the rise of domestic debt markets may deepen financial systems, foster local investor bases, and enhance monetary autonomy, the authors warn that the line between financial deepening and financial repression can be thin.

Editor’s note: This article was first published on VoxEU.

Over the past two decades, Africa’s public debt landscape has undergone a profound transformation. The continent’s total public debt has risen more than fourfold since the early 2000s, reaching $2 trillion in 2024. But just as important as the increase in debt volumes is the shift in structure. While the story of African debt has often been told in terms of Eurobonds (Hooper et al. 2022), Chinese loans (Jiang 2025), and multilateral financing, the real revolution has been happening closer to home in domestic debt markets.

Until recently, analysts lacked comprehensive evidence to assess how African governments borrow domestically. Existing datasets were either aggregate, self-reported, or limited to external debt (Mihaly and Trebesch 2023). The result is a striking asymmetry: while the composition of Africa’s external liabilities was relatively well understood, domestic debt remained largely invisible.

To fill this gap, we created the African Debt Database (ADD), a new open-access, instrument-level dataset covering more than 50,000 loans and securities issued by 54 African countries between 2000 and 2024. The database, built from tens of thousands of primary government documents, allows for the first time a detailed view of both domestic and external debt. It includes information on currency, maturity, interest rate, creditor type, and issuance terms. In Manger et al. (2025) we describe how the dataset was constructed and presents four key findings that emerge from its first exploration.

Building the African Debt Database

Constructing a continent-wide sovereign debt dataset was a massive data-collection effort that took several years and combined manual curation, automated parsing, and machine-learning tools. The project drew on hundreds of national sources to capture every recorded government borrowing instrument from 2000 to 2024.

Data sources and scope

The ADD covers 54 countries. For each instrument – whether a bond, bill, or loan – the dataset records issuance and maturity dates, coupon structure, interest rate, currency, amount issued, and creditor information. It distinguishes between:

• domestic debt, defined by currency of denomination, encompassing treasury bills and government bonds issued in local markets; and
• external debt, comprising foreign-currency loans and bonds contracted under international law.

Domestic data were retrieved from more than 60 separate official websites and archives. In many cases, historical records had disappeared from public view and we reconstructed them using the Internet Archive’s Wayback Machine to recover missing auction calendars and bulletins. The data were then standardised into a unified format to allow comparability across countries and over time.

Digitisation and validation

A key innovation of the ADD is its digitisation pipeline. Thousands of PDF and scanned documents were processed using a combination of optical character recognition (OCR) and rule-based extraction scripts in Python. When the source material was too unstructured for automated parsing, we used large language models to extract table-like information from text. This hybrid approach proved particularly effective for semi-structured bond listings and auction announcements.

Each record was manually verified and, where possible, cross-checked against commercial data providers such as Bloomberg and Refinitiv to ensure accuracy. Missing yields or prices were inferred using standard fixed-income formulas, and all values were converted into US dollars at the exchange rate prevailing at issuance.

In total, the dataset encompasses roughly $6.3 trillion in nominal commitments, making it the most comprehensive repository of African sovereign debt ever assembled. All code and documentation are freely available at www.africandebtdatabase.com, allowing other researchers to replicate or update the analysis.

What the new data reveals

This new dataset allows researchers and policymakers to look behind the aggregates and trace how African governments actually borrow. The early evidence highlights four structural facts that are reshaping the continent’s debt landscape.

1. The domestic debt boom

Since 2010, the volume of domestic debt issuance has tripled, rising from about $150 billion to nearly $500 billion. On average, African governments now raise more than half of their financing at home, reversing decades of dependence on external lenders.

The early phase of this boom was dominated by short-term Treasury bills. Over time, however, several middle-income economies have built functioning local bond markets with regular issuance calendars and longer-dated maturities.

This shift represents a major structural change. Borrowing in domestic currency reduces exposure to exchange-rate depreciation and the original sin (Eichengreen et al. 2023, Hausmann and Panizza 2010, Onen et al. 2025) of dollar-denominated liabilities. Yet it also transfers risk inward, concentrating sovereign exposure in local financial institutions. Domestic banks and pension funds, which now hold a large share of government securities, are increasingly entangled with fiscal dynamics – a classic form of sovereign-bank nexus (Dunz et al, 2024) that merits close monitoring.

2. Divergent borrowing costs

The dataset provides the first continent-wide comparison of interest rates across all debt instruments. Multilateral loans remain by far the cheapest source of finance, averaging below 1%, followed by bilateral official lending. In contrast, domestic bonds and bills are the most expensive, carrying nominal yields of 10–13% on average.

When adjusted for inflation, however, many of these domestic instruments produce negative real yields. The heterogeneity across countries is striking. Nations such as Tanzania and Mauritius issue at modest real rates, whereas Nigeria and Egypt have seen real yields fluctuate between –30% and +10%. Several factors could explain this divergence: varying degrees of financial repression, differences in monetary credibility, and structural limits to investor choice in thin domestic markets. For researchers, these patterns open a new field of inquiry into how inflation expectations, capital controls, and domestic savings constraints shape sovereign borrowing costs in low-income economies.

3. Maturities all over the map

The maturity structure of African debt is highly uneven. At one end of the spectrum, South Africa and Egypt issue ten- to fifteen-year domestic bonds and maintain relatively deep yield curves. Countries like Tanzania, Uganda, and Mauritius are gradually lengthening maturities, supported by improved macroeconomic stability and institutional reforms. At the other end, countries such as Ghana and Mozambique are locked in a cycle of short-term borrowing.

4. Refinancing risks rising

The new dataset also allows a granular look at repayment profiles. Debt-service obligations are projected to exceed $100 billion in 2026 and remain above $60 billion per year through 2030. Most of these payments arise from bonds, both domestic and international, which carry higher costs and shorter maturities than concessional loans.

Countries with large Eurobond exposure face particularly steep repayment cliffs. Even those that have avoided default spend between 20% and 40% of fiscal revenue servicing debt. As access to external markets tightens, many governments are turning to domestic issuance to roll over maturing liabilities. The result is a self-reinforcing debt treadmill: short-term obligations financed by even shorter-term instruments.

This dynamic underscores a growing vulnerability. Without deeper domestic markets or renewed access to concessional finance, refinancing risks could quickly morph into liquidity crises especially if domestically issued debt is held by foreign investors (Carstens and Shin 2019). For international financial institutions, the challenge is to support transparent local-currency market development while guarding against the build-up of rollover and interest rate risks.

Policy implications and the road ahead

Taken together, these findings point to a historic shift.

On the positive side, the rise of domestic debt markets may deepen financial systems, foster local investor bases, and enhance monetary autonomy. Governments able to borrow in their own currency are less exposed to exchange-rate crises and may gain greater control over fiscal policy.

But the risks are equally clear. Large domestic debt stocks can crowd out private credit, heighten the links between sovereign and banking-sector stability, and create incentives for inflationary finance when fiscal pressures intensify. The line between financial deepening and financial repression can be thin.

A second and equally important lesson from the ADD is that transparency is feasible. Despite limited funding and no institutional backing, a small academic team was able to compile a detailed, verifiable record of Africa’s debt instruments by combining public data sources with modern digital tools. If such an effort can succeed on a shoestring budget, the potential for well-resourced international institutions to replicate it at scale is considerable.

Better data are not just a research tool – they are a prerequisite for better policy. As Africa’s borrowing landscape evolves, credible, granular, and timely information will be essential for designing debt-sustainability frameworks, developing domestic markets, and coordinating among creditors.

References

Carstens, A and H S Shin (2019), “Emerging Markets Aren’t Out of the Woods Yet How They Can Manage the Risks”, Foreign Affairs, 15 March.

Dunz, N, E Feyen, C Mousset and F Stewart (2024), “The rise of sovereign–bank nexus risks in developing economies”, VoxEU.org, 16 October.

Eichengreen, B, R Hausmann and U Panizza (2023), “Yet it Endures: The Persistence of Original Sin”, *Open Economies Review *34(1): 1-42.

Hausmann, R and U Panizza (2010), “Redemption or abstinence?”, VoxEU.org, 21 February.

Hooper, E, V Le Clainche and C Seitz (2022), “Debt in Sub-Saharan Africa”, Tresor Economics, No. 299, January.

Jiang, Z (2025), “The rise of China as an international lender”, VoxEU.org, 25 April.

Manger, M, D Mihalyi, U Panizza, N Rescia, C Trebesch and K Wong (2025), “Africa’s Domestic Debt Boom: Evidence from the African Debt Database”, CEPR Discussion Paper No. 20747.

Mihalyi, D and C Trebesch (2023), “Who Lends to Africa and How”, Kiel Working Paper No. 2217

Onen, M, H S Shin and G von Peter (2025), “Overcoming original sin: Shedding new light on uneven progress”, Economic Policy 40: 575–620.