Over the past two decades, African sovereign marketable debt has expanded substantially, largely supported by favourable macro-financial conditions and strong demand for Emerging Markets and Developing Economies (EMDE) debt. However, since 2022, the financing landscape has become markedly more challenging, characterised by rising borrowing costs, weakening foreign demand and heightened geopolitical and trade-related uncertainties. At the same time, access to official lending has declined, prompting many African countries to rely more heavily on costlier market-based financing. These shifts, coupled with large refinancing needs, risk pushing debt servicing costs beyond the fiscal capacity of some sovereigns. In this context, strengthening debt management frameworks and progressively developing of local currency bond markets will be essential to support fiscal sustainability and financial resilience.
Africa Capital Markets Report 2025
Abstract
This chapter examines Africa’s sovereign debt markets, including issuance trends, borrowing costs, foreign market access, and credit rating developments. The analyses predominantly use available market data on central government bond issuance, with methodological details in the Annex.
Sovereign issuers in Africa face diverse financing needs and constraints. Unlike most developed countries, where borrowing is largely covered through local currency bonds, many African economies rely on a mix of local and foreign currency bonds, bilateral loans, and concessional finance.
As of the end of 2024, only 60% of African countries had sovereign bonds outstanding, the lowest share across regions. This likely reflects the high number of low-income, small economies rather than explicit policy decisions.
Between 2007 and 2024, Africa’s annual sovereign debt issuance rose from USD 70 to 350 billion, and outstanding bond debt stock grew from USD 160 to USD 730 billion. Issuances as a share of GDP tripled from 5% to 15%, the highest among EMDEs, yet Africa still represents only 1% of global sovereign bonds despite accounting for 3% of global GDP.
About 60% of Africa’s marketable debt is in fixed-rate local currency bonds. However, this share is mainly driven by a few large issuers, as the median share of fixed-rate debt across African countries is 40%, the lowest among regions, with higher use of Treasury bills and foreign currency debt, and nearly no use of variable-rate debt.
The average term to maturity (ATM) of outstanding debt varies from as low as two years to over ten years. Larger and more frequent issuers such as Kenya, Morocco, Nigeria and South Africa generally maintain longer ATMs.
Africa has the lowest share, at 25%, of marketable relative to total government debt across regions. Larger economies and those issuing mainly local currency bonds generally have higher marketable debt. African countries also issue across more bond lines and use fewer buybacks and re-openings than advanced economies, making liquid benchmark lines harder to establish.
About 80% of rated African countries were classified as high risk or below in 2024, the highest share across regions. Only Botswana and Mauritius have investment-grade ratings. Côte d’Ivoire and Seychelles are, since 2007, the only countries upgraded to speculative grade to maintain this rating. As such, around half of Africa’s outstanding debt is high risk, and the share of investment-grade debt is negligible compared to other regions.
In 2024, Africa’s real yields on local currency bonds reached 5%, their highest level since at least 2007 and well above their real effective interest rates of around 1%, as a result of lower costs on official lending. Thus, refinancing all debt at current market rates without altering debt-to-GDP ratio trajectories would require primary balances to increase by an average of 2.5% of GDP, severely straining fiscal space.
USD-denominated bond yields in Africa reached 9% in 2024, compared to 7% in 2007, and are now the highest among all regions. The number of African countries issuing in foreign markets fell sharply in 2022 and 2023. Despite a rebound in 2024, net foreign borrowing has remained negative every year since 2022.
Over the past two decades, African sovereign marketable debt has grown significantly, driven largely by favourable macro-financial conditions and strong investor demand for EMDE debt. Unlike most developed economies, where borrowing is predominantly through local-currency bonds, many African countries rely on a combination of local- and foreign-currency bonds, bilateral loans, and concessional financing.
By the end of 2024, around 60% of the 54 African countries analysed had sovereign bonds outstanding (Figure 5.1, Panel A).1 Of these, 7 had sovereign bonds outstanding only in foreign markets, 12 only in local markets, and 12 in both markets. For the remaining 23, no bond data was available in LSEG, implying that they are not sovereign issuers. Eastern and Middle Africa, sub-regions where low-income countries are more prevalent, had the lowest share of sovereign issuers, both at 50%, while Southern Africa had the highest, at 80%.
Africa has the lowest share of sovereign issuers among EMDEs, reflecting the correlation between sovereign issuance and a country’s economic size and per capita income (Figure 5.1, Panels B and C). All EMDEs with a GDP above USD 1 trillion are sovereign issuers, and more than 90% of those with a GDP between USD 300 billion and USD 1 trillion are as well. However, only about 60% of countries with GDP below USD 300 billion issue sovereign bonds. In Africa, over 95% of EMDEs fall into this low-GDP group as of 2024. Income levels also play a role: about 85% of high-, upper-middle-, and lower-middle-income EMDEs are sovereign issuers, but this drops to around 60% among low-income countries. Africa again stands out, with the highest share of low-income countries at roughly 35%, compared to just 2.5% outside the region.
Although Africa has the largest number of low-income countries and the lowest share of sovereign issuers among EMDEs, it accounts for the largest share of issuers among low- and lower-middle-income countries. This means that, when controlling for income, Africa has more sovereign issuers than other regions. As a result, the continent’s overall low share of issuers reflects structural constraints rather than policy choices.
Note: Panel A shows all countries that had a sovereign bond outstanding at the end of 2024 and maturing in the following years. Countries that are not shown are not issuers, or their data is not available. Panel B reports this share for 2024, broken down by region and income group. Panel C displays, for each year, the share of EMDEs issuing a bond out of the total in the region.
Source: LSEG and OECD calculations.
Sovereign debt issuance and debt levels rose significantly in Africa between 2007 and 2024. Annual issuance increased from USD 70 billion to USD 350 billion, while the outstanding marketable debt stock rose from USD 160 billion to USD 730 billion (Figure 5.2, Panels A and B). This growth represented the second-highest annual increase across regions during the period (Figure 5.3, Panel A). As a result, annual issuance as a share of the sovereign issuer’s GDP in Africa moved from being the lowest across regions in 2007 to the highest in 2024 (Figure 5.3, Panel B).
Source: LSEG and OECD calculations.
The increase was particularly marked over 2011-2021, when both annual issuance and the outstanding debt-to-GDP ratio doubled in Africa. Several factors contributed to this surge, which also affected EMDEs in other regions. First, market access improved on the back of stronger fundamentals, supported by high foreign exchange reserves built during the commodity boom in the early 2010s. This was especially the case for African economies reliant on exports of oil, minerals, and agricultural products (Afreximbank, 2025[1]). Second, improvements in monetary policy credibility, driven by the wider adoption of inflation-targeting frameworks and floating exchange rates, also enabled greater market access (Kalemli-Ozcan and Unsal, 2024[2]). Third, accommodative monetary policy in advanced economies, characterised by low interest rates and quantitative easing, led some investors to move into EMDE debt to secure positive real returns, thus increasing the investor base. Finally, external shocks, notably the global financial crisis and the pandemic, increased fiscal needs (AfDB, 2025[3]).
Whilst the level of debt spiked across the region as whole, the distribution is more even between countries than in other regions. In 2024, the top 5 and top 10 issuers accounted respectively for 79% and 91% of the region’s total outstanding debt, compared to 86% and 97% respectively in other regions (Figure 5.3, Panel C). This reflects not only the smaller number of large economies in Africa but also the presence of many issuers with meaningful levels of outstanding debt.
Note: Regions’ figures include only EMDEs.
Source: LSEG and OECD calculations.
The majority (around 60%) of Africa’s marketable debt is denominated in fixed-rate local currency bonds (Figure 5.4, Panels A and B). Fixed-rate local currency bonds carry the lowest risk for the issuer, especially when issued with long maturities. They are crucial for developing the local risk-free yield curve, which is foundational for local currency bond markets, and makes debt transparency a salient issue for the development of local financial markets (Box 5.1). The median share of fixed-rate debt in Africa, an indicator more representative of most countries in the region, is around 80%, the highest across regions. Amongst four of the larger issuers—Algeria, Kenya, Morocco and South Africa—fixed-rate debt accounts for 75% or more of the total, while Egypt, the second largest issuer overall in Africa, is something of an outlier with fixed-rate debt accounting for less than 25% of the total.
Africa’s relatively low share of variable-rate debt is also notable. Inflation-linked bonds can be useful for extending the yield curve and developing specific market segments. They are particularly valuable for countries with a history of high inflation or for capturing demand from market segments involving investors with inflation-linked liabilities, such as pension funds and insurance companies. Because investors pay a premium for inflation protection, inflation-linked bonds can be cost-effective in both advanced economies (OECD, 2024[4]) and EMDEs (Cardozo and Christensen, 2024[5]). While unforeseen increases in inflation lead to higher interest payments for inflation-linked bonds, this is typically hedged to some degree by simultaneously increasing tax receipts related to the higher inflation. In addition to advanced economies, some EMDEs, especially in the Latin America and the Caribbean region, where countries have been relying heavily on these instruments in their financing programmes.
Note: Variable rate includes floating rates and inflation-linked bonds. Regions’ figures include only EMDEs.
Source: LSEG and OECD calculations.
The weighted average term-to-maturity (ATM) of Africa’s outstanding marketable debt is 7.4 years—around one year shorter than in Asia and half a year shorter than the Latin America and the Caribbean region (Figure 5.5, Panel A). This figure has increased slightly from around 6.7 years in 2016. The rise mainly reflects increased issuance of foreign currency debt, which typically has longer maturities than local currency debt, while the ATM for both types of debt has remained relatively stable since 2016: around 6.5 to 6.9 years for local currency debt and 10 to 11 years for foreign currency debt (Figure 5.5, Panel C).
These regional averages mask significant variation in the ATM among African sovereign issuers. Some countries have ATMs as low as 2 years, while others exceed 10 years (Figure 5.5, Panel B). Relatively larger and more frequent issuers such as Kenya, Morocco, Nigeria and South Africa generally have longer ATMs in both local and foreign currency markets. Two exceptions are Algeria and Egypt. Although they are among the largest issuers in Africa, their ATMs for local currency debt remain low—around 3 years for Algeria and under 2 years for Egypt. Egypt’s ATM has been below 2 years since 2017 and has not exceeded 3 years since 2007. Algeria’s ATM was around 8 years in 2007, declined until 2016, and has stayed below 4 years since. In contrast, many smaller issuers operate in only one market—either in local or foreign currency—but their debt maturities vary widely.
Note: Regions’ figures include only EMDEs. LCY refers to local currency and FCY refers to foreign currency.
Source: LSEG and OECD calculations.
The accurate and timely reporting of data on outstanding debt and issuance activities by issuers is critical to support investor confidence and improve market access. It also ensures potential debt vulnerabilities can be identified early, reduces information asymmetries, and makes it easier to hold officials accountable. This box is based on publicly available data from official sources of 15 selected African economies and complements existing efforts to improve debt data transparency. It focuses on local currency marketable debt, and is complemented by Annex 5.B, which provides country-specific findings, with general findings summarised in this box. Overall, debt data is well reported, but transparency in key areas could still be improved.
Information on the composition of the outstanding debt is generally available in official sources. All countries analysed provide a breakdown of debt into external and domestic debt, although not all are clear about the definitions used to classify debt into these categories (e.g. investor residence or currency). All issuers also disclose the instrument composition of the local marketable debt (e.g. bill and bond), but about half do not provide a breakdown of non-marketable debt (e.g. loans, pensions, arrears), hindering a more thorough analysis.
Primary market data on auction calendars and results is typically readily accessible but can be outdated. The delay in the publication of auction results varies, but generally they are published on the same day as the auction. In some cases, auction calendars are not up to date, with only about half of the countries maintaining calendars of upcoming issuances for this year.
All countries publish some form of a debt report, with most having a separate Medium-Term Debt Strategy (MTDS) document. Typically, domestic debt service for previous years is also provided in the debt report. However, information on the investor base is only published by about one-third of the countries analysed.
Availability of recent outstanding debt data is limited in about half of the countries (Figure 5.6, Panel A). The ease of accessibility of this data also differs across countries. Some countries have multiple periods for which data is inaccessible or in formats that cannot be easily compared. Additionally, a few countries have inconsistencies between different data sources. For example, one country reports different values for total outstanding bonds in different publications for the same year.
Further inconsistencies exist when comparing official sources and two third-party sources, Bloomberg and LSEG (Figure 5.6, Panel B). Just over one third of the countries analysed show a difference of six percentage points or less between two sources, allowing for an accurate understanding of the true debt burden. However, for another third, the difference between the market and official sources is above 25 percentage points of the total debt. Given the importance of data reliability in public debt, in these cases, market data providers and debt management offices could coordinate to resolve data inconsistency.
Note: 2025 data in Panel A data comes from any period until June 2025. Panel B uses the most recent data available from government sources, as shown in Panel A, except for Angola, where data is limited to 2019 from LSEG. Missing data indicates no recent data in the source. Calculations use total debt data from the IMF in the denominator.
Sources: G7 and Paris Club (2023), “Preliminary Findings from the G-7 and Paris Club Countries Debt Data Sharing Exercise”, https://thedocs.worldbank.org/en/doc/6e72b0ded996306fa01f5db7a0c38b19-0050052021/related/G7-and-Paris-Club-Data-Reconciliation-Exercise-April-2023.pdf; World Bank. (2024, October). Debt Reporting Heat Map - 2024. Retrieved from https://www.worldbank.org/en/topic/debt/brief/debt-transparency-report/2024.
A key distinction between EMDEs and advanced economies lies in how public borrowing is structured. EMDEs rely more on non-market sources, such as loans from multilateral institutions, bilateral creditors, and private banks. In contrast, advanced economies align financing needs closely with market-based bond issuance, with debt management funding virtually all borrowing needs in markets. In many EMDEs, however, the decision to issue bonds involves not only consideration of borrowing needs, but also the availability of other funding sources, such as loans, and is therefore a policy choice rather than the standard funding mechanism.
Although non-market borrowing tends to be cheaper, it offers less flexibility and financial autonomy. Between 2000 and 2020, in Africa, the average interest rate on foreign currency bonds was 6%, compared to 3% on loans from Chinese banks and 1% from multilateral lenders (Mihalyi and Trebesch, 2023[6]). However, loans alone are often insufficient to meet sovereign financing needs, as they can be limited in size, slow to structure, and carry currency risk. Concessional loans can also be earmarked for certain types of expenditures or projects, reducing flexibility in the use of proceeds.
Development of a local bond market entails creation of an extensive informational, legal, and institutional infrastructure. Once a robust and safe infrastructure is established and markets are developed, governments can issue larger volumes over time by accessing a broader creditor base when compared to the limited number of loan providers. Market instruments also typically carry no or fewer restrictions on the use of proceeds compared to official lending, enhancing fiscal flexibility. Additionally, local currency bonds, specifically, help mitigate external shocks by reducing currency and maturity mismatches (Eichengreen and Hausmann, 1999[7]).
It is important for sovereign issuers to develop fungible instruments that can serve as liquid benchmarks, helping to reduce liquidity premiums and, in turn, lower funding costs. Creating several benchmark bonds across the yield curve also promotes domestic market development by providing a risk-free yield curve essential for price discovery for private sector securities. As economies grow and public financing needs increase, reliance on bond markets tends to increase.
Africa, with its large number of low- and lower-middle-income countries and smaller economies, has the lowest share of marketable debt relative to total government debt across regions (Figure 5.7, Panel A). This share has generally increased over time, peaking in 2022 at more than one quarter, before declining since amid a deterioration in global market conditions.2 Upper-middle-income countries in Africa typically have a higher marketable debt share than countries from other regions in the same income group, while the opposite is true for lower-middle-income African countries (Figure 5.7, Panel B). Unlike in Asia, upper-middle-income countries in Africa show a much higher share of marketable debt than lower-middle-income countries.
Source: LSEG, IMF World Economic Outlook April 2025, BIS Debt Security Statistics, OECD Financial Accounts, and OECD calculations.
Two variables strongly correlated with the share of marketable debt in Africa are economic size (GDP) and the proportion of local currency bonds in total marketable debt (Figure 5.8, Panel A). Larger economies are generally better positioned to issue debt in their own currency abroad, broadening the investor base and reducing sovereign-bank nexus risks (Eichengreen, Hausmann and Panizza, 2022[8]).3 Mauritius and Namibia are notable exceptions: despite relatively small economies, both maintain marketable debt shares above 60%. Moreover, no African country with a local currency bond share below 25% of total marketable debt has marketable debt exceeding 20% of total government debt. Conversely, in all cases where marketable debt surpasses 60% of the debt, local currency bonds make up over 75% of marketable debt, showing the importance of local currency markets for financing capacity.
The government’s capacity to raise revenues also plays a critical role in developing sovereign bond markets. Countries with high reliance on Official Development Assistance (ODA) generally have lower marketable debt shares; all countries whose ODA represented over 5% of GDP in 2024 had marketable debt shares around or below 40% (Figure 5.8, Panel B). In contrast, countries with marketable debt shares above half of total government debt all had tax-to-GDP ratios exceeding 15% (Figure 5.8, Panel C), highlighting the link between tax capacity and market access.
Note: Panels are as of 2024 or 2023, depending on data availability. LCY refers to local currency debt.
Source: LSEG, IMF World Economic Outlook April 2025, BIS Debt Security Statistics, OECD Financial Accounts, OECD Official Development Assistance dataset, OECD Revenue Statistics, and OECD calculations.
Credit ratings are another key determinant in the development of bond markets, serving as a reference point and often a precondition for international investors' decisions. Several institutional investors, in particular, can only invest in securities with ratings above a specified threshold4. As such, credit ratings directly influence both the size of the investor pool accessible to a sovereign issuer and the cost of borrowing. Securing a favourable rating from one of the three major agencies can therefore be a pivotal moment for sovereigns seeking market-based financing.
Most African countries face significant challenges in this regard, as they either have high credit risk, remain unrated, or are not active sovereign issuers (Figure 5.9, Panel A). By the end of 2024, only Botswana and Mauritius held investment-grade ratings in Africa, while Côte d’Ivoire, Morocco, Seychelles and South Africa were rated as speculative grade.5 The remaining 24 rated African sovereigns were assessed as high risk or default equivalent.6 Consequently, they face higher borrowing costs and limited access to capital from risk-averse institutional investors, which is also captured by their shares in some bond indexes (see Box 5.2).
Bond indices are benchmarks that track the performance of a selection of bonds with a common theme or characteristics and allow for passive investments into a broad range of bonds. Furthermore, they serve as benchmarks for actively managed investment mandates. Bond indices therefore play an important role by creating structural demand for the constituents’ debt instruments. Common indices that give investors exposure to emerging market sovereign debt include the JP Morgan Emerging Market Bond Global Diversified Index (JPM EMBIGD), JP Morgan Government Bond Index Emerging Market (GBI-EM) Global Diversified, MSCI Emerging Markets Sovereign Bond Index (MESBI), and the FTSE Emerging Markets Government Bond Index (FTSE EMGBI).
While the requirements to be included in each of these indices vary, generally, these indices include sovereign emerging market debt instruments that meet minimum size and liquidity requirements: typically USD 500 million per security and a residual maturity of over one year, accessibility to foreign investors and a credit rating between a certain range (usually higher than default but lower than AAA). Some indices include only local currency debt, while others include only debt issued in USD.
A country’s weight in an index usually depends on its total share of eligible outstanding debt. As such, countries with a larger marketable debt stock will often have a higher share in the index. For indices like JPM EMBI and JPM GBI-EM, country weights are also capped, ensuring more diversified exposure.
The share of African debt in all these indices remains below 15%. The FTSE EMBI allocates about 12.5% to Central and Eastern Europe, Middle East and Africa (CEEMA). While the JPM GBI-EM Global Diversified (local market debt) includes only South Africa, making the total weight for the continent just 7%, the JPM EMBI Global Diversified (USD-denominated debt) includes 15 African countries, for a total weight of only 12%.
To increase their share in these indices, where possible, African countries can focus on developing liquid benchmarks by issuing bonds that reach the minimum size for inclusion. Additionally, where applicable, the removal of capital controls would allow their debt to be more easily accessible to foreign investors, especially since countries have been excluded from indices in the past due to foreign exchange convertibility issues. Countries should also focus on credit ratings, as defaulted bonds are excluded.
Source: FTSE Russell (2025[13]), FTSE Emerging Markets Government Bond Index (EMGBI) Factsheet; MSCI (2025[14]), MSCI Emerging Markets Sovereign Bond Index (MESBI), https://www.msci.com/documents/1296102/182d7ae4-0f37-23c4-ffca-57a6fe57eaa6; JP Morgan (2023[15]), J.P. Morgan EMBI Global Diversified Index, https://www.jpmorgan.com/content/dam/jpm/cib/complex/content/markets/composition-docs/gbi-em-gd-factsheet.pdf
This large share of high-risk countries in Africa has not changed significantly over the past 15 years, with Africa consistently having the highest shares of these countries when compared to other regions (Figure 5.9, Panels B and C). Since the commodity boom in the mid-2010s, credit quality has deteriorated in Africa, with high-risk countries representing nearly 80% of the rated countries at the end of 2024. In other regions, it fluctuated around half of the rated countries. The only two African countries that were upgraded from high risk to speculative and that have retained that rating until now are Côte d’Ivoire (in 2015) and Seychelles (2015). All other substantial rating upgrades since 2007 have subsequently been downgraded. Egypt (2011), Gabon (2016) and Nigeria (2015) were downgraded from speculative grade to high risk, while Namibia (2017), Tunisia (2012) and South Africa (2017) lost their investment grade rating in the past 15 years. Namibia (2022) and Tunisia (2013) were further downgraded to high-risk countries.
Note: Panel A does not show countries without a credit rating and countries for which the credit rating is not available in the dataset considered. Panel C includes only countries with credit ratings below high risk in the share. Regions’ figures include only EMDEs.
Source: LSEG and OECD calculations.
The fact that Africa is the region with the highest share of low- and lower-middle income countries does not completely explain its higher share of high-risk countries. In fact, more than 85% of Africa’s lower-middle income countries have a credit rating equivalent to high risk or below by end 2024, against less than 65% for countries in the same income group in other regions (Figure 5.10, Panel A). As a result, about half of all the outstanding debt from Africa is high risk, compared to 10% in Latin America and the Caribbean and a negligible amount in Emerging and Developing Asia (Figure 5.10, Panel B). Additionally, in other regions, nearly half or more than half of all the outstanding debt was issued by investment-grade countries. In Africa, this share is negligible.
Note: UMIC* excludes China. Regions’ figures include only EMDEs.
Source: LSEG and OECD calculations.
A number of reasons have been put forward to explain the disparities in credit ratings and consequently in the cost of borrowing between African countries and other regions. One explanation highlights biases in the global financial architecture, including as a result of regulatory reforms following the Global Financial Crisis that made credit rating agencies more cautious, leading to conservative ratings that may not fully reflect country fundamentals (Cash and Khan, 2024[9]; Gwaindepi, 2025[10]). Other reasons include low transparency in the budget process, the high significance of the informal sector, limited financial development, weak public institutions, political risks, and the contagion effect of debt defaults and economic distress among African countries (Gbohoui, Ouedraogo and Some, 2023[11]; AfDB, 2025[3]; Alter et al., 2025[12]). Regardless of the causes, the impact of credit ratings weighs heavily on African countries’ borrowing costs and impacts their access to funding (see Box 5.2).
The cost dynamics of local and foreign currency debt differ significantly. Local currency sovereign bond yields set the benchmark risk-free rate, incorporating inflation, real returns, term premiums, and limited credit risk. When most debt is denominated in local currency, its cost is a crucial indicator of solvency, as liquidity risks are minimal given the sovereign’s ability to issue currency. Real interest rates drive borrowing costs and fiscal sustainability. This means that even if nominal yields are high, negative real rates (when inflation outpaces nominal yields) can allow debt-to-GDP ratios to remain stable, despite primary deficits and elevated nominal borrowing costs. In contrast, foreign currency bonds usually form a small share of sovereign debt, where the main concern is liquidity risk—ensuring the country can obtain the foreign currency to service its obligations. Yields on these bonds include a spread over the yield curve of the currency-issuing country, reflecting the premium investors require to absorb this risk.
In 2024, Africa’s estimated real yields on local currency bonds reached their highest level since at least 2007, at nearly 5% (Figure 5.11, Panel A). While this remains below the real rates in Latin America and the Caribbean, it exceeds those in Emerging Asia, which were around 2.5% in 2024. Moreover, several countries are characterised by a higher level of local currency yields compared to forecasted growth in the next five years, pointing to possible future fiscal challenges (IMF, 2025[13]). High-risk and speculative grade countries generally face higher real yields (Figure 5.11, Panel C), and given Africa’s high share of such countries, it is notable that real yields remain lower than in Latin America and the Caribbean. One likely reason is that Latin America and the Caribbean have experienced more frequent hyperinflation episodes since the 1980s, whereas in Africa, these events have been rarer and typically confined to smaller economies like Zimbabwe. As a result, Africa has benefited from relatively lower real borrowing costs in local currency debt, given their credit risk profile. This underscores the importance of monetary policy credibility in determining sovereign borrowing costs.
Note: 1) The figures comprise a sample of bonds with primary market yield to maturity available in LSEG. All values are averages weighted by issuance amounts in the respective instrument by country. Real market interest rates are estimations as they can only be precisely computed for matured bonds. These were estimated by adjusting nominal yields from a sample of primary market issuance for realised and projected inflation from 2007 to 2030. Bonds excluded from Panel A are those maturing after 2030, with no available yield-to-maturity at issuance data in LSEG and from countries with recent events of hyperinflation. 2) Regions’ figures include only EMDEs. 3) Panel B estimates the primary balance that would stabilise debt-to-GDP ratios using the growth and inflation rates of the latest available year on the IMF’s forecasts (2030), and the implied effective interest rate of the 2024 debt, as reported by the IMF. 4) Panel C include s EMDEs from all regions.
Source: LSEG, OECD Revenue Statistics, IMF World Economic Outlook April 2025, and OECD calculations.
The development of local currency bond markets can strengthen fiscal flexibility and buffer external shocks, but progress must be gradual due to their higher costs relative to the current costs of Africa’s debt. Bilateral and concessional lending has helped keep real effective interest rates near zero or negative in at least seven African sovereign issuers, about one-third of the sample in this analysis (Figure 5.11, Panel B). The average real effective interest rate in Africa is around 1%, well below the nearly 5% average of local currency real bond yields. Only South Africa has real market rates meaningfully below its real effective rates because most of its debt is already marketable. If all debt were refinanced at market rates, African countries would need to raise their primary balances by an average of 2.5 percentage points of GDP to stabilise debt-to-GDP ratios, severely limiting fiscal space for investments and government expenditures, given that African countries have about 15% of tax-to-GDP ratios. In the most extreme case, Zimbabwe re-issuing its debt at market rates could reduce their fiscal space by nearly 10 percentage points of GDP.
Yields at issuance for USD-denominated bonds in Africa also reached record highs in 2024, nearing 9%, with four countries, Angola, Cameroon, Kenya and Nigeria issuing bonds with yields above 10% (Figure 5.12, Panels A and B). This is the highest among all regions, with weighted averages elsewhere remaining below 7% that same year. 2024 is not an exception—since 2010, Africa has consistently paid higher yields on these bonds, averaging around 7% in 2014-2023, compared to roughly 5% or lower for EMDEs in other regions. These elevated costs cannot be explained by income levels alone. For example, lower-middle-income African countries faced yields roughly one percentage point higher than their peers elsewhere for foreign currency bonds issued between 2022 and 2024 (Figure 5.12, Panel C). A major factor behind these higher borrowing costs is Africa’s large share of high-risk countries, as credit risk plays a critical role in foreign currency debt markets. Developing local currency bond markets according to internationally recognised guidelines can help in this respect (Hashimoto et al., 2021[14]).
Source: LSEG and OECD calculations.
One impact of the recent rise in yields was that many African countries decided not to or were unable to issue in foreign markets in 2022-2023 (Figure 5.13, Panels A and B). Eurobonds are often issued in large volumes to ensure sufficient market liquidity, but meeting these volumes at high costs can be prohibitive for some sovereigns. Countries also tend to follow a less consistent Eurobond issuance schedule when compared to their local currency bond issuances, with some issuers opting to tap the market only when conditions are more favourable, occasionally going years without issuing at all. Thus, as few as four African countries issued in foreign markets in 2022 (Angola, Egypt, Nigeria and South Africa), and just three in 2023 (Côte d'Ivoire, Egypt and Morocco), compared to ten in 2021. This represented the steepest decline across all regions and resulted in negative net borrowing in foreign markets across Africa in consecutive years, the first time this has happened since before 2007 (Figure 5.13, Panel C).7
In 2024, a brief decline in USD yields led some African sovereigns to front-load foreign market issuances as an important source of foreign currency, but this was insufficient to make net borrowing in foreign markets positive. Although 2024 saw the highest number of African countries issuing in foreign markets since 2007, net foreign borrowing remained negative. These issuances primarily offset redemptions rather than providing new financing. In contrast, net issuance in local currency markets declined but remained positive, highlighting its importance in supporting fiscal autonomy, flexibility, and reduced external vulnerability.
The partial cessation of foreign market issuance in 2022–2023 also affected countries that predominantly issue in euros. While USD bonds are often preferred for their liquidity and deep investor base, some countries issue predominantly in euros, such as Benin, Côte d’Ivoire and Senegal. These countries have traditionally preferred issuing in euros to mitigate currency mismatch risks, given the euro currency peg in the West African CFA Franc currency union. However, even these issuers halted foreign bond issuance during 2022–23, underscoring the sharp decline in demand for EMDE debt among investors from advanced economies.
Source: LSEG and OECD calculations.
Although yields on foreign currency debt are typically lower than those on local currency debt in African countries, the local currency costs of foreign currency debt often exceed those of local currency debt due to currency depreciation (Figure 5.14, Panel A).8 Specifically, a breakdown of local currency costs since 2007 shows that while the yield component averages around 7% annually, an average depreciation of 6% in local currencies pushed the total cost of foreign currency bonds above 10% for every year from 2007 to 2021 (Figure 5.14, Panel B). In 2023 and 2024, the depreciation effect was minimal, likely because these bonds were issued recently and currency depreciation typically occurs gradually. Indeed, between 2007 and 2024, African currencies have gradually depreciated against the US dollar, with a total depreciation of 65% in these years (Figure 5.14, Panel C).
Note: All panels refer to African countries only, unless a regional breakdown of EMDE countries is provided. The costs of foreign currency debt in local currency are estimated using a product increment formula that multiplies the annual yield at issuance of the bond by the annualised depreciation of the local currency against the bond’s currency of denomination. For bonds that have not yet matured, the foreign exchange rate at the end of 2024 is assumed to remain constant until maturity. The currency index refers to a cross-country average weighted by 2024’s GDP.
Source: LSEG and OECD calculations.
Compared to other regions, Africa is unique in that the median carry costs for foreign currency bonds, when converted to local currency, exceeded those for local currency bonds in 2024 (Figure 5.15, Panel A). This is because African foreign currency bond yields are the highest among EMDEs, even though the depreciation effects are similar to those in Latin America and the Caribbean (Figure 5.14, Panels C and Figure 5.15, Panel C). Africa stands out for the high nominal costs of its foreign currency debt, even as its local currency real yields remain below those of Latin America and the Caribbean. This makes local currency borrowing generally cheaper and less risky for most African countries when compared to foreign currency borrowing from markets.
In 2024, foreign currency carry costs exceeded those of local currency debt in 13 of the 16 African countries with data available for both markets (Figure 5.15, Panel B). The exceptions were Benin, which issues mainly in euros and has a currency pegged to the euro; Kenya, which halted issuing foreign bonds in 2021 before the recent surge in yields; and Tanzania, whose last foreign currency bond issuance was in 2013 when yields were much lower. In some cases, annual foreign currency carry costs in local currency terms were substantially higher, reaching up to 25% (e.g. Ghana). These patterns indicate that foreign currency bonds in Africa have not only generally been more expensive but also riskier, as currency depreciation under stress scenarios can trigger sharp cost spikes.
Despite these challenges, there are still incentives for African countries to issue foreign currency bonds. They provide benchmarks for private sector foreign currency borrowing, can supply foreign exchange in the absence of concessional loans, and typically offer lower immediate costs since currency depreciation occurs over time and initial nominal rates are below local currency yields. This dynamic can create a temporary boost to fiscal space, which can also be appealing from a political economy perspective.
Note: The costs of foreign currency debt in local currency are estimated using a product increment formula that multiplies the annual yield at issuance of the bond by the annualised depreciation of the local currency against the bond’s currency of denomination. For bonds that have not yet matured, the foreign exchange rate at the end of 2024 is assumed to remain constant until maturity. Carry costs are estimated using the same approach but are based on the stock of bonds outstanding at the end of each year.
Source: LSEG and OECD calculations.
Sovereign debt management is intrinsically connected to the development of government securities markets, through the choice and design of instruments, issuance strategies, consistent market communication, and a commitment to transparency. Debt management offices help develop an investor base for sovereign debt, including both local and foreign investors, and increase market liquidity by providing benchmark securities that serve as collateral and as a safe asset for investors. They can also broaden the investor base by promoting sovereign retail programmes (Box 5.1), which can help mobilise domestic savings and strengthen market resilience. As a result of a diversified investor base and liquid securities, sovereign issuers can stimulate a continuous demand for their securities, helping them obtain funding even in times of stress, and attain liquidity premiums in their bond issuances, reducing the costs paid on interest.
One of the main tools employed in debt management to enhance market liquidity and secure liquidity premiums is the development of liquid benchmarks across the yield curve. Benchmarks are bond lines built to a certain size at key maturities, enabling large investors to trade without significantly affecting prices. This reduces transaction costs and meets the needs of large investors. Beyond providing liquidity premiums and demand for sovereign securities, benchmarks aid price discovery for private borrowers who price their securities above the sovereign curve. Without established benchmarks in local currency, the private sector may resort to foreign currency borrowing, increasing the economy's vulnerability to global shocks and currency mismatches. In times of distress, this can feed back to the government, affecting economic activity, tax revenues, and even sovereign risk.9
Note: The methodology to count the number of benchmarks can be found at Annex 5.C. Short-term is defined as below 7 years, medium term from 7 to 13 years, and long-term above 13 years.
Source: LSEG and OECD calculations.
Some African countries issue across more bond lines than advanced economies, which can hinder the establishment of key benchmarks with substantial volumes (Figure 5.16). Specifically, the median number of benchmark bonds in advanced economies is two up to the 7-year tenor (short segment of the yield curve), one between 7 and 14 years (medium segment), and one above 14 years (long segment). In comparison, African countries show higher medians of three, two, and two, respectively. This pattern is influenced by a few countries that issue across a broader range of bond lines, with some—such as Namibia, South Africa, and Tanzania—having four or more lines with significant volumes in the medium or long segments.
Establishing benchmarks involves more than concentrating issuances in key bond lines. Large benchmarks force issuers to manage redemption schedules with sizeable principal repayments. This often involves using tools such as buybacks of securities before maturity to reduce the cash needed for these payments, typically conducted in the year of redemption (OECD, 2025[15]). Buybacks can also support liquidity in bond lines that no longer serve as benchmarks or remove them entirely from circulation, particularly when combined with new bond issuances (switch operations) (OECD, 2023[16]). In addition, building large bond lines often requires multiple re-openings, as reaching the desired volume in a single issuance would be impractical and costly.
A smaller share of African issuers conducts bond buybacks and re-openings account for a lower share of borrowing in Africa when compared to EMDEs in other regions (Figure 5.17, Panels A and B). Only about 33% of African sovereign issuers carried out buybacks of bonds maturing within one year (often linked to cash management) in 2024, the lowest rate, together with Asian issuers. Among bonds with maturities exceeding one year, Africa also has the lowest share of sovereign issuers conducting buybacks, at just 17%, resulting from a significant decline in such operations in 2024. Similarly, since 2015, the share of marketable borrowing from re-openings has been the lowest in Africa when compared to other regions (Figure 5.17, Panel C), with re-openings in more than half of African sovereign issuers accounting for around or less than 20% of market borrowing in 2024. This is much lower compared to other regions, where borrowing from markets via re-openings often makes up at least half of the total (Figure 5.17, Panel D).
Source: LSEG and OECD calculations.
Another way to foster liquid markets is by having a regular and predictable issuance schedule. This involves not only publishing issuance calendars in advance, but also conducting similar-sized auctions for the same bond lines and maintaining a consistent frequency of issuance for certain bonds. Such practices allow investors to better plan and manage their portfolios, which can boost participation and improve pricing in primary markets. Moreover, predictability reduces uncertainty and enhances market confidence, supporting the development of a deeper and more liquid secondary market over time. While this approach may be more feasible for issuers with reliable market access, others may need to rely to some extent on opportunistic issuance, depending on market conditions. Still, efforts to adopt more predictable practices can, over time, help broaden and stabilise the investor base, ultimately contributing to more reliable market access in a reinforcing feedback loop.
African countries generally exhibit greater variation in auction sizes for the same bond lines compared to EMDEs from other regions, while maintaining a similar auction frequency (Figure 5.18, Panels A and C). Compared to advanced economies, all African sovereign issuers except Morocco and Uganda show more variability in auction sizes (Figure 5.18, Panel B). Variations in auction size are not necessarily problematic if they align with auction plans. However, unexpected changes can increase uncertainty and reduce market confidence, harming liquidity. If they reflect opportunistic issuance, such as selling more bonds in favourable conditions and fewer in unfavourable ones, investors may anticipate these shifts, undermining demand and liquidity.
Note: Annex 5.D. includes country-specific auction charts contrasting selected advanced and African economies’ issuance patterns.
Source: LSEG and OECD calculations.
Another driver of auction size variation is limited cash management capacity. When cash management is weak, issuances must align more closely with immediate cash needs, leading to unpredictable auction sizes or schedules. Therefore, the ability to produce reliable forecasts of government cash flows is essential for predictable and transparent issuance. A well-developed cash management function enables governments to separate bond issuance from short-term cash shortfalls, supporting adherence to the issuance calendar and improving predictability for investors.
Although sovereign issuers are in a strong position to influence the development of sovereign bond markets through their primary market operations, several other stakeholders' policies and actions are also extremely relevant. This includes especially the central bank, financial market regulators, and providers of trading, payment, clearing, and settlement systems—these are essential for establishing the pre-conditions for market development, such as a credible macroeconomic environment with low inflation, stable fiscal policies, and sufficient savings; an active money market with central bank support and transparent regulation; robust secondary market liquidity backed by a sufficient volume of securities and reliable intermediaries; and strong market infrastructure and legal frameworks to protect investors and ensure market integrity (Hashimoto et al., 2021[14]). Therefore, the development of sovereign bond markets is ultimately a multi-faceted effort that requires political commitment and sustained coordination across government and financial market participants.
To facilitate retail access to sovereign debt, many sovereign issuers adopt specific retail programmes. These often feature dedicated products, digital platforms and specific communication campaigns. They are adopted because retail investors may be less aware of the opportunity to invest in sovereign debt and less willing or able to buy government bonds on the secondary market. Direct participation is also perceived as a more effective way to reach households in markets where there are limited options for retail investors to access collective investment schemes and other similar vehicles. Several African countries have adopted retail programmes and/or products, with Kenya having among the highest investor participation globally.
Africa lags behind other regions in formal banking access, with particularly low level of bank account ownership in Central and West Africa. This would limit governments' ability to distribute bonds through traditional banking channels. However, the region is among the highest users of digital and mobile finance globally, presenting a significant opportunity for sovereign issuers to adopt more inclusive, technology-enabled public financing strategies (OECD, 2024[17]). In this context, retail bonds delivered through digital financial services and fintech innovations can serve as a valuable complement to the financing programme while supporting financial education. Kenya stands out for significant retail investor participation. The Kenyan government has actively promoted retail investment in its debt instruments through initiatives like the M-Akiba mobile bond platform. Launched in 2017, M-Akiba allows individuals to invest in government bonds via mobile phones with a minimum investment of KSH 3 000 (approximately USD 30), aiming to enhance financial inclusion. As of June 2024, retail investors in Kenya held approximately 17% of the country's domestic sovereign debt. This represents a significant increase from 13% in June 2023, indicating a growing participation of retail investors in government securities.
In South Africa, the RSA Retail Savings Bonds programme, launched in 2004, offers fixed and inflation-linked bonds to individual investors with a minimum investment of ZAR 1 000 (approximately USD 20). Despite their availability, the uptake among retail investors remains modest, with retail investors accounting for approximately 0.3% of the country's domestic sovereign debt holdings. South Africa is currently in the process of redesigning its sovereign retail programme, as part of the National Treasury’s broader digital transformation agenda.
In Nigeria, the FGN Savings Bond is an investment vehicle that allows individual investors to lend funds to the Federal Government of Nigeria. In 2025, two and three-year fixed rate bonds are available with a minimum investment of NIG 5000 (approximately USD 3) and a maximum of NIG 50 000 000 (approximately USD 30 000). As of September 2024, retail investors in Nigeria held approximately 0.1% of the country's domestic sovereign debt.
Meanwhile, Egypt is actively exploring the issuance of sovereign retail debt products. The Ministry of Finance has announced plans to launch retail bonds in 2025 as part of its strategy to diversify public debt instruments and enhance financial inclusion.
Sources: Foxall & Policino, 2025, ‘’Sovereign retail debt programmes and instruments: A review of country practises”, https://www.oecd.org/en/publications/sovereign-retail-debt-programmes-and-instruments_e2a782d0-en.html; FSD Africa, 2024, ‘’The story of Kenya’s m-akiba: selling treasury bonds via mobile’’, https://fsdafrica.org/blog/the-story-of-kenyas-m-akiba-selling-treasury-bonds-via-mobile/; The National Treasury, Republic of Kenya, 2024, ‘’Annual public debt report 2023-24’’, https://www.treasury.go.ke/wp-content/uploads/2024/11/Annual-Public-Debt-Management-Report-.pdf; Government of South Africa, 2025, ‘’RSA Retail Savings Bond’’, https://www.rsaretailbonds.gov.za/Home.aspx; Government of South Africa, 2025, ‘’Treasury invites bids for Structured Debt Advisory to support RSA Retail Bonds Programme’’, https://www.gov.za/news/media-statements/treasury-invites-bids-structured-debt-advisory-support-rsa-retail-bonds.
The substantial increase in Africa’s sovereign marketable debt since 2007 occurred during a period of much more favourable macro-financial conditions. Specifically, Africa’s sovereign bond markets have expanded significantly, with the marketable debt-to-GDP ratio rising from 13% in 2007 to 30% in 2024. Much of this growth happened in the past decade when global interest rates were low and a commodity boom supported demand for EMDE debt. These conditions enabled central banks to build foreign reserves and strengthen monetary policy tools, reducing debt service costs. However, since 2022, macro-financial conditions have deteriorated significantly: borrowing costs have surged to multi-decade highs in advanced economies and EMDEs, and uncertainties have risen due to geopolitical conflicts and trade disruptions. In parallel, borrowing in OECD countries is at record highs in 2024 and is expected to increase further in 2025 (OECD, 2025[18]), exceeding even the COVID-19 borrowing peak and reducing demand for EMDE securities.
At the same time, official creditors—traditionally a cheaper funding source—are reducing their lending and grants to EMDEs. ODA from OECD countries, critical for many African countries either directly or through multilateral loans, is declining, with an estimated 7.1% drop in 2024 and further projections of a 9% to 17% fall in 2025 based on OECD surveys (OECD, 2025[19]). China, which has provided below-market-rate loans to Africa (Mihalyi and Trebesch, 2023[6]), has also scaled back lending in recent years (Afreximbank, 2025[1]). Furthermore, as some countries graduate from low- to middle-income status, they lose access to some concessional loans; since 2007, this has affected Benin, Côte d'Ivoire, Ghana, Kenya, Nigeria, Senegal, Tanzania, Zambia and Zimbabwe.
The combination of large refinancing needs due to increased debt and reduced official assistance will likely push African countries to rely more on markets, which could substantially raise debt servicing costs. If all debt had to be refinanced at current market rates, countries would need to raise primary balances by an average of 2.5% of GDP to prevent debt-to-GDP ratio trajectories from worsening, given that current average real effective interest rates of around 1% are far below local market real yields near 5%. Foreign currency debt is even costlier, reflecting high yields since 80% of Africa’s rated sovereign issuers are classified as high-risk or below, compared to about half for EMDEs in other regions.
This challenging environment underscores the need to develop deeper local bond markets, which can mitigate financial vulnerabilities, enhance the effectiveness of monetary policy, and encourage investment. Achieving this, however, requires a long-term strategy and close coordination across fiscal, monetary, and regulatory policies. In this context, prudent debt management and a gradual shift toward greater reliance on local currency bond markets are critical. As official lending declines, African sovereigns may increasingly rely on markets, making it essential to strengthen tax capacity to support higher debt costs. At the same time, maximising the use of concessional borrowing remains important, as its lower cost frees up fiscal space for growth- and development-oriented expenditures.
Sovereign issuers could also explore options to enhance liquidity, reduce costs and diversify their investor base. This could include issuing inflation-linked bonds, widely used by some EMDEs to reduce costs if inflation stays on target; concentrating more issuances in benchmark bond lines; increasing the use of buybacks and re-openings; adopting more stable auction sizes; and exploring retail debt programmes. Finally, coordination with central banks, financial market regulators, and providers of trading, payment, clearing, and settlement systems is also crucial to strengthen sovereign bond market foundations and expand the local investor base, especially as foreign investors absorb record-high domestic bond supply amid high fiscal needs and quantitative tightening (OECD, 2025[18]). This is a challenging environment, and decisive policy action is essential to strengthen debt sustainability and financial resilience.
References
[3] AfDB (2025), “Africa Financing: Stability Mechanism”, Technical and Operational Report, https://www.afdb.org/en/documents/african-financing-stability-mechanism-afsm-technical-and-operational-report.
[1] Afreximbank (2025), “African Debt Outlook: A Ray of Optimism”, Market Updated Volume 3, https://www.afreximbank.com/reports/african-debt-outlook-a-ray-of-optimism/.
[12] Alter, A. et al. (2025), “Navigating the Evolving Landscape of External Financing in Sub-Saharan Africa”, IMF Working Paper No. 2025/139, https://doi.org/10.5089/9798229016100.001.
[5] Cardozo, C. and J. Christensen (2024), “The Benefit of Inflation-Indexed Debt: Evidence from an Emerging Bond Market”, Federal Reserve Bank of San Francisco Working Paper 2023-04, https://doi.org/10.24148/wp2023-04.
[9] Cash, D. and M. Khan (2024), “Rating the globe: reforming credit rating agencies for an equitable financial architecture”, United Nations University: Centre for Policy Research, https://unu.edu/cpr/report/rating-globe-reforming-credit-rating-agencies-equitable-financial-architecture.
[7] Eichengreen, B. and R. Hausmann (1999), “Exchange Rates and Financial Fragility”, NBER Working Paper No. 7418, https://doi.org/10.3386/w7418.
[8] Eichengreen, B., R. Hausmann and U. Panizza (2022), “Yet it Endures: The Persistence of Original Sin”, Open Economies Review (2023) 34:1–42, https://doi.org/10.1007/s11079-022-09704-3.
[20] Escolano, J. (2010), “A Practical Guide to Public Debt Dynamics, Fiscal Sustainability, and Cyclical Adjustment of Budgetary Aggregates”, https://www.imf.org/external/pubs/ft/tnm/2010/tnm1002.pdf.
[21] FTSE Russell (2025), FTSE Emerging Markets Government Bond Index (EMGBI) Factsheet.
[11] Gbohoui, W., R. Ouedraogo and Y. Some (2023), “Sub-Saharan Africa’s Risk Perception Premium: In the Search of Missing Factors”, IMF Working Paper WP/23/130, https://www.imf.org/en/Publications/WP/Issues/2023/06/23/Sub-Saharan-Africas-Risk-Perception-Premium-In-the-Search-of-Missing-Factors-534885.
[10] Gwaindepi, A. (2025), “How can the global sovereign debt architecture work better for African countries?”, Danish Institute for International Studies (DIIS), DIIS Report Vol. 2025 No. 03, https://research.diis.dk/portal/en/publications/how-can-the-global-sovereign-debt-architecture-work-better-for-african-countries(735e60fb-c75d-4ad1-b2b5-ad540f1e93b4).html.
[14] Hashimoto, H. et al. (2021), “Developing Government Local Currency Bond Markets”, International Monetary Fund and The World Bank Guidance Note, https://documents1.worldbank.org/curated/en/790921615526044752/pdf/Guidance-Note-for-Developing-Government-Local-Currency-Bond-Markets.pdf.
[13] IMF (2025), Global Financial Stability Report - October 2025, https://www.imf.org/en/Publications/GFSR/Issues/2025/10/14/global-financial-stability-report-october-2025.
[23] J.P. Morgan (2024), The Government Bond–Emerging Market Index Series, https://www.jpmorgan.com/content/dam/jpm/cib/complex/content/markets/composition-docs/gbi-em-gd-factsheet.pdf.
[22] J.P. Morgan (2023), J.P. Morgan EMBI Global Diversified Index, https://www.jpmorgan.com/content/dam/jpm/cib/complex/content/markets/composition-docs/pdf-27.pdf.
[2] Kalemli-Ozcan, S. and F. Unsal (2024), “Global Transmission of FED Hikes: The Role of Policy Credibility and Balance Sheets”, NBER Working Paper No. w32329, https://ssrn.com/abstract=4794379.
[6] Mihalyi, D. and C. Trebesch (2023), “Who lends to Africa and how? Introducing the Africa Debt Database”, Kiel Working Paper, No. 2217, https://ideas.repec.org/p/zbw/ifwkwp/2217.html.
[24] MSCI (2025), MSCI Emerging Markets Sovereign Bond Index, https://www.msci.com/documents/1296102/182d7ae4-0f37-23c4-ffca-57a6fe57eaa6.
[18] OECD (2025), Global Debt Report 2025: Financing Growth in a Challenging Debt Market Environment, OECD Publishing, Paris, https://doi.org/10.1787/8ee42b13-en.
[15] OECD (2025), Managing Government Cash: A Review of Practices in OECD Countries, OECD Publishing, Paris, https://doi.org/10.1787/7675eb58-en.
[19] OECD (2025), “Preliminary official development assistance levels in 2024”, Detailed Summary Note, https://one.oecd.org/document/DCD(2025)6/en/pdf.
[17] OECD (2024), Fintech Lending in Sub-Saharan Africa, https://www.oecd.org/content/dam/oecd/en/topics/policy-sub-issues/digital-finance/FinTech-lending-in-Sub-Saharan-Africa.pdf.
[4] OECD (2024), Global Debt Report 2024: Bond Markets in a High-Debt Environment, OECD Publishing, Paris, https://doi.org/10.1787/91844ea2-en.
[16] OECD (2023), “OECD Sovereign Borrowing Outlook 2023”, OECD Publishing, Paris, https://doi.org/10.1787/09b4cfba-en.
Primary sovereign bond market data are based on original OECD calculations using data obtained from the London Stock Exchange Group (LSEG), which provides international security-level data on new issues of sovereign bonds. The data set covers bonds issued by emerging market sovereigns in the period from 1 January 2000 to 31 December 2024 and includes both short-term and long-term debt. Short-term debt (“bills”) is defined as any security with a maturity of less than or equal to 365 days but no less than 33 days, as bill issuance with a maturity of less than 33 days or less is done for cash management purposes and is excluded from calculations. Bonds issued by central banks that have non-budgetary financing purposes were excluded. The data provides detailed information for each bond issue, including the proceeds, maturity date, interest rate, and currency structure.
This report's definition of emerging markets is consistent with the IMF’s classification of Emerging Market and Developing Economies (EMDEs) used in its World Economic Outlook. The regional definitions are also those used by the IMF, while the income categories used (high income, low income, lower middle income, upper middle income) are those from the World Bank as of 2024, which is based on GNI per capita levels. EMDE sovereign bond issuers are:
High income countries (HICs): Bahamas (LAC), Bahrain (MECA), Barbados (LAC), Bulgaria (EMDEs Europe), Chile (LAC), Hungary (EMDEs Europe), Kuwait (MECA), Oman (MECA), Panama (LAC), Poland (EMDEs Europe), Qatar (MECA), Romania (EMDEs Europe), Russian Federation (EMDEs Europe), Saudi Arabia (MECA), Seychelles (SSA), Trinidad and Tobago (LAC), United Arab Emirates (MECA), Uruguay (LAC).
Upper middle income countries (UMICs): Albania (EMDEs Europe), Algeria (MECA), Argentina (LAC), Armenia (MECA), Azerbaijan (MECA), Belarus (EMDEs Europe), Belize (LAC), Bosnia Herzegovina (EMDEs Europe), Botswana (SSA), Brazil (LAC), China (EMDEs Asia), Colombia (LAC), Costa Rica (LAC), Dominican Republic (LAC), Ecuador (LAC), El Salvador (LAC), Equatorial Guinea (SSA), Fiji (EMDEs Asia), Gabon (SSA), Georgia (MECA), Grenada (LAC), Guatemala (LAC), Indonesia (EMDEs Asia), Iraq (MECA), Jamaica (LAC), Kazakhstan (MECA), Malaysia (EMDEs Asia), Maldives (EMDEs Asia), Mauritius (SSA), Mexico (LAC), Moldova (EMDE Europe), Mongolia (EMDEs Asia), Montenegro (EMDEs Europe), Namibia (SSA), North Macedonia (EMDEs Europe), Paraguay (LAC), Peru (LAC), Serbia (EMDEs Europe), South Africa (SSA), Suriname (LAC), Thailand (EMDEs Asia), Türkiye (EMDEs Europe), Ukraine (EMDEs Europe), Venezuela (LAC).
Lower middle income countries (LMICs): Angola (SSA), Bangladesh (EMDEs Asia), Benin (SSA), Bolivia (LAC), Cameroon (SSA), Cote Ivoire (SSA), Egypt (MECA), Ghana (SSA), Honduras (LAC), India (EMDEs Asia), Jordan (MECA), Kenya (SSA), Lao PDR (EMDEs Asia), Lebanon (MECA), Morocco (MECA), Myanmar (EMDEs Asia), Nicaragua (LAC), Nigeria (SSA), Pakistan (MECA), Papua New Guinea (EMDEs Asia), Philippines (EMDEs Asia), Senegal (SSA), Sri Lanka (EMDEs Asia), Swaziland (SSA), Tanzania (SSA), Tunisia (MECA), Uzbekistan (MECA), Viet Nam (EMDEs Asia), Zambia (SSA).
Low-income countries (LICs): Burkina Faso (SSA), Chad (SSA), Congo (SSA), Ethiopia (SSA), Guinea Bissau (SSA), Malawi (SSA), Mali (SSA), Mozambique (SSA), Niger (SSA), Rwanda (SSA), Togo (SSA), Uganda (SSA), Yemen (MECA).
The geographical classification of countries in Figure 5.1 and Figure 5.9 is aligned with United Nations publication "Standard Country or Area Codes for Statistical Use":
Northern Africa: Algeria, Egypt, Morocco, Tunisia.
Eastern Africa: Ethiopia, Kenya, Malawi, Mauritius, Mozambique, Rwanda, Seychelles, Tanzania, Uganda, Zambia.
Western Africa: Benin, Burkina Faso, Côte d’Ivoire, Ghana, Guinea-Bissau, Mali, Niger, Nigeria, Senegal, Togo.
Middle Africa: Angola, Cameroon, Gabon, Republic of the Congo.
Southern Africa: Botswana, Eswatini, Namibia, South Africa.
A number of bonds have been subject to reopening. For these bonds, the initial data only provides the total amount (original issuance plus reopening). To retrieve the issuance amount for such reopened bonds, specific data on the outstanding amount on each reopening date for the concerned bonds have been downloaded separately from LSEG. As the reopening data only provides amounts outstanding, the outstanding amount on the previous date is subtracted from the outstanding amount on that given date to obtain the issuance amount on each relevant date. These calculated issuance amounts are converted on the transaction date using USD foreign exchange data from LSEG. To ensure consistency and comparability, the same method is used for all bonds, including those not subject to reopening.
LSEG provides information on the market of issuance of each bond. When this variable is available, the market of issuance is classified as “Domestic” if LSEG classifies it as “Domestic” or “Domestic (others)” and as “Foreign” in all other cases.
Outstanding debts in local currency are converted to USD using end-of-year foreign exchange rates. Exchange offers and certain bonds in the dataset have been manually excluded when they did not have a Bond ID identifier (ISIN, RIC or CUSIP) and when they could not be manually confirmed by comparing them with official government data.
LSEG provides rating information from three leading rating agencies: Fitch, Moody’s, and S&P. For each country with rating information in the dataset, a value of 1 is assigned to the lowest credit quality rating (C or below) and 21 to the highest credit quality rating (AAA for Fitch and S&P, and Aaa for Moody’s). Non-investment grade categories include ratings up to BB+ for Fitch and S&P, and up to Ba1 for Moody’s.
The rating in question is then assigned to each relevant bond issued by that country (as at issuance or transaction date). If ratings are available from several agencies, their average is used. Final ratings are categorised as follows: those equal to or higher than 15 are classified as Investment Grade A (IG A); ratings falling between 12 and 14 are designated as Investment Grade BBB (IG BBB); ratings between 9 and 11 are categorised as Speculative Grade BB (SG BB); and ratings below 9 are classified as Single B high risk (Single B and below). Within the high-risk category, ratings equal to or lower than 3.5 indicate a default or very high risk of default.
When computing the number of upgrades and downgrades, ratings data are observed on a monthly basis, excluding those equal to 1. If a country has received several ratings from the same agency in one month, the latest one is used. The weighted debt quality analysis uses rating information from three rating agencies (Fitch, Moody’s and S&P). The rating valid at the end of the year for a country is assigned to the totality of its outstanding debt stock. The share is then computed as a stock-weighted average across rating groups.
The computation of primary market yields relies on LSEG data regarding the prices and yields of each issuance. These data are available only for a limited number of bonds.
Yields and prices are assigned to our dataset based on the Bond ID and its date of issuance. When multiple prices or yields are available for the same bond on the same date of issuance, the average value is considered.
When a yield or a price has the same Bond ID but not the same date of issuance, we consider the price or yield as having the closest issuance date for those Bond IDs falling within a five-day range before or after, beginning with one day later, followed by one day before, and so on.
When yields are not readily available, they are computed using the corresponding price and the R package called jrvFinance. Prices below zero or higher than 999 are excluded. In addition, yields that exceed 100 or fall below -5 are excluded. Yields that are more than 10 times the median annual yields for each issuer, considering the same instrument type and maturity category, are also excluded.
The debt stabilising primary balance is computed following the methodology outlined by Escolano (2010[20]) and using data from the IMF World Economic Outlook as of April 2025 for the latest year available (2030) for all variables except the real effective interest rate, which used data as of 2024 or the estimated market real yield.
Where pb is the general government's primary balance as a ratio of GDP; d is the general government debt stock as a ratio of GDP; ip is the real effective interest rate, and; g is the real GDP growth.
Panel A: Local currency marketable debt refers to the sum of bonds and bills that are issued in local currency. Years are shaded for which data for local currency marketable debt values can easily be accessed from an official website (e.g. Ministry of Finance, Treasury or Central Bank), and understood to be referring to bills and bonds.
Panel B: This panel uses data from the most recent year in which information from official sources is available, as shown in Panel A. The share of local currency marketable debt is calculated by dividing the total local currency marketable debt from each source (official, Bloomberg and LSEG) by the IMF total outstanding debt for that year.
It is possible that governments record securities as bills and bonds, but that they are not tradeable. This could be one reason why official data tends to be higher than Bloomberg and LSEG.
The data for this table comes from official government websites, usually the Treasury, Ministry of Finance or Central Bank. Where data is not readily available or easily accessible on these websites, a "-" is used.
The column “Marketable breakdown” refers to the breakdown of marketable outstanding debt into its securities (i.e. bonds and bills). The column “Non-marketable breakdown” refers to the breakdown on non-marketable outstanding debt into its securities or obligations (e.g. loans, arrears).
The column "Most recent auction calendar" shows the most recent year for which an auction calendar can be found on an official website. If information relating to upcoming auctions is published ad hoc or no regular calendar is readily available, a "-" is used. Similarly, "Most recent auction results" shows the most recent year for which an auction calendar can be found on an official website.
The column “Auction results: price / yield” shows whether auction results including yield and or price are published.
The column "Most recent annual debt report" shows the most recent year for which an annual report can be found on an official website. Where monthly or quarterly reports are comprehensive, the most recent year of the report is used.
The column “Medium-Term Debt Strategy” shows the number of upcoming years that are covered by the strategy while column "Most Recent MTDS" shows the most recent year for which a Medium-Term Debt Strategy can be found on an official website.
The "Language" column shows the language in which debt related documents are available. It does not include the language in which websites are available.
The following links are the primary links that were used:
Table Notes
1 Refers to an official government department responsible for debt management, like the Ministry of Finance or Treasury. For this table, this excludes the Central Bank.
2 The official securities exchange in Portugal
3 Banque des États de l'Afrique Centrale
4 Regional agency supporting issuance and management of public debt in West African Economic and Monetary Union (WAEMU)
5 Direction General de la Dette Publique (Debt Management Office)
6 Central Bank of Egypt
7 National Bank of Ethiopia
8 Central Bank of Kenya
9 Debt Management Office
10 South African Reserve Bank
11 Banque Centrale de Tunisie
12 Bank of Uganda
A quantitative method was used to count the significant and distinct issuance peaks by time to maturity, providing a measure of how concentrated (benchmark-focused) or diffuse each country's sovereign bond issuance is.
Step 1 – Kernel Density Estimation
For each country, a smoothed distribution of issuances by time to maturity was estimated using a weighted kernel density function:
Where:
is time to maturity (in years)
is the weight for each issuance
is the smoothing parameter (bandwidth adjuster, set to 0.1)
Step 2 – Peak Detection
We identify local maxima (peaks) in the estimated density using the standard second-derivative method:
All such peaks were then retained only if their satisfied the relative prominence condition:
Where:
s the relative height threshold (e.g. 10% of the maximum density)
Step 3 – Distinct Peak Filtering
To ensure we count only distinct peaks (not small bumps clustered near each other), we apply a minimum spacing rule:
Where:
years is the minimum required horizontal distance between peaks
When two peaks are closer than , only the taller peak is retained
Step 4 – Classification by Maturity Band
Each peak is assigned a maturity band: Short-term: ; Medium-term: ; Long-term:
These peaks can be identified visually, as in the figures below.
Source: LSEG and OECD calculations. Note: The values on the x-axis represent the years to maturity at issuance. The values on the y-axis represent the empirical probability density and not the empirical probability itself. The area under the density curve across the x-axis equals 1. The density is weighted by the issuance volumes.
Source: LSEG and OECD calculations. Note: The values on the x-axis represent the years to maturity at issuance. The values on the y-axis represent the empirical probability density and not the empirical probability itself. The area under the density curve across the x-axis equals 1. The density is weighted by the issuance volumes.
Source: LSEG and OECD calculations. Y-axis refers to tenor, size of bubble is proportionate to auction amount.
Source: LSEG and OECD calculations. Y-axis refers to tenor, size of bubble is proportionate to auction amount.
← 1. This analysis is based on the data availability in LSEG.
← 2. Worsening financial conditions might decline the share of marketable debt through two means: first, by causing countries to halt bond issuances or shifting the instrument composition of their borrowing to cheaper instruments; and second by reducing the mark-to-market value of bonds in the debt portfolio.
← 3. Over-reliance on local banks as investors for sovereign bonds can also pose risks due to the sovereign bank nexus. When domestic banks hold large amounts of domestic government debt, and, thus, the public and financial sectors are closely linked, the stress in one can destabilise the other. A drop in sovereign bond values weakens the financial institutions that hold them, increasing financial stability risks. If these institutions require government support, the resulting fiscal strain can further undermine confidence in public finances, negatively affecting bond prices and creating a self-reinforcing cycle.
← 4. These thresholds for eligible investment typically vary from BBB- to BBB+, depending on the investor.
← 5. High risk refers to a credit rating of single B or lower, while very-high risks or in default to CCC- or lower.
← 6. At the end of 2024, 30 African countries had been rated by at least one of the three major credit rating agencies. South Africa was the first, receiving a rating in 1994. By 2000, six African countries had a credit rating, increasing to 20 by 2006 and reaching 30 by 2019.
← 7. Net issuance (i.e. the amount issued minus the amount redeemed within a specific period) is a proxy for the marginal exposure investors take. While this net figure does not fully capture the issuers' reliance on foreign investors, since some will buy their bonds directly in the local rather than international market, it does offer a useful gauge of foreign demand.
← 8. The cost of foreign currency-denominated debt in EMDEs includes two main components: nominal yield and currency effect. Yields on EMDEs’ local currency debt are generally higher than on foreign currency-denominated debt, as a currency risk premium is priced in higher macro-economic instability and inflation in EMDEs. However, since the USD has tended to appreciate against EMDE currencies over time, coupon and principal payments typically become costlier when converted to local currencies (the currency effect).
← 9. This explains why the concept of Original Sin is assessed by considering not only the sovereign but also the entire country’s assets and liabilities exposure (Eichengreen and Hausmann, 1999[7]).