SEVEN GATES RESEARCH · MACRO · AFRICA
GDP Does Not Pay the Coupon
The IMF and World Bank have changed the way they assess sovereign debt. The useful part for investors is not another threshold. It is learning to see the strain before it reaches the headline ratio.
By The Lokoja Contrarian · 25 September 2026 · Seven Gates Research
At a glance
| What changed First major overhaul of the joint IMF-World Bank low-income-country debt framework since 2017. | Operational Expected from the second half of 2027. |
| What gets more attention Domestic debt, gross financing needs, interest burden and the financial system carrying the debt. | Countries Nigeria, Egypt, Ghana, Kenya, South Africa, Zambia. |
| Carrying capacity Nigeria 53, Egypt 46, Ghana 61, Kenya 54, South Africa 74, Zambia 54. | Where investors look Maturity, auctions, FX, sovereign-heavy banks and companies crowded out by government borrowing. |
At four in the morning on 28 March 1979, a pump stopped in the Unit 2 reactor at Three Mile Island, on the Susquehanna River in Pennsylvania. A relief valve opened to release pressure, as it was designed to do. Then it stuck open.
On the control panel, a light went out. The light told the operators that the valve had been ordered to close. It said nothing about whether the valve had obeyed. Within minutes more than a hundred alarms were sounding. The men in the room were trained, awake and certain. Another gauge showed the water level in the pressuriser climbing, and they read it the way their training taught them to read it: a system with too much water in it. So they throttled back the emergency cooling.
The reactor was losing coolant through the open valve. The rising level was steam and water surging towards the leak. For more than two hours the plant drained while the instruments in front of the operators told a coherent, reassuring and wrong story. By the time a shift supervisor closed a second valve behind the stuck one, part of the core had melted.
The plant did not fail for want of instruments. It had hundreds of them. It failed because one reading was trusted to speak for the whole system, and every other reading was interpreted so that it agreed.
Sovereign-debt analysis has its own indicator light. It sits in the middle of every country presentation, every rating report and every newspaper chart: public debt as a share of GDP. It is a real instrument. It measures something that matters. The trouble is the mood it sets in the room.
A low reading feels like a quiet night shift. At 36 per cent the needle rests in the green, the coffee is warm and nobody walks the floor. Nobody asks how much of this year's revenue has already been promised to creditors, how many Treasury bills fall due before Christmas, who is holding them, or what those holders have stopped doing for everyone else so they can keep holding them. A high reading produces the opposite reflex. At 80 per cent the alarms sound, even when the debt sits long-dated, in local currency, with patient pension funds, inside a market deep enough to absorb it.
So the ratio can be right about the level and wrong about the danger, in both directions at once. Some sovereigns at 80 per cent sleep well. Some at 35 per cent are already in trouble, and the light on the panel is still saying the valve is shut.
The answer is not a better single number. It is an inspection. You walk the plant and put your hand on the valve itself, rather than trusting the lamp that reports the order. For a sovereign that means three plain questions. Where does the cash come from? When does it have to go out? And who stands in between, holding the paper and deciding whom else to lend to?
Those questions matter most on one particular day.
Coupon day has no interest in your economy.
The date was fixed when the bond was sold, often years earlier. On that date the bondholder asks for an amount, in a named currency, paid into a named account, by close of business. The finance ministry cannot pay it in GDP. GDP belongs to farmers, traders, refiners and hairdressers. The ministry can pay only with money it has collected or money it can borrow again.
That distinction sounds too obvious to need saying. Sovereign-debt analysis has spent fifty years forgetting it.
Walter Wriston, who ran Citicorp through the 1970s, liked to reassure people that countries don't go bankrupt. In a narrow legal sense he was right. In August 1982 Mexico told its creditors it could not pay. Mexico was still there the next morning. So was the debt. So, for most of the following decade, was the problem.
Debt-to-GDP earned its place on the panel because it is simple, broadly comparable and gives some sense of the claim creditors hold over an economy. The trouble starts when a useful ratio acquires ambitions. In 2010 Carmen Reinhart and Kenneth Rogoff published a paper suggesting that growth weakened sharply once public debt passed 90 per cent of GDP. Finance ministers in several capitals quoted it. Three years later a graduate student at the University of Massachusetts, Thomas Herndon, tried to replicate the result for a class assignment and found, among other problems, a spreadsheet error. The cliff looked a good deal gentler after the audit. The habit of looking for a cliff survived.
Nigeria shows why the habit misleads.
The IMF estimates Nigerian public gross debt at 36.1 per cent of GDP in 2025. By the standards of that 90 per cent debate, Nigeria hardly needs discussing. The same IMF tables put Federal Government interest payments at 53.2 per cent of FGN revenue. Consolidated government revenue is only 10.2 per cent of GDP.
Nothing here is inconsistent. Nigeria has a large economy. The Federal Government has a rather smaller wallet. One figure measures the size of the house. The other measures the salary of the person paying the mortgage.
That gap between the economy and the wallet is a good way into the first major overhaul of the IMF-World Bank low-income-country debt framework since 2017.
Six sovereigns, 529 million people
About $1.59tn of nominal GDP in 2026. Same continent; very different financing machinery.
Figure 1. The six economies together represent about 529 million people and $1.59 trillion of projected 2026 nominal GDP. Source: IMF country pages for population; IMF World Economic Outlook, April 2026, for nominal GDP and GDP per capita; Seven Gates Research.
The old sovereign crisis followed a script anyone in emerging markets could recite. A government borrowed in dollars. Reserves thinned. The exchange rate became excitable. Creditors began returning calls rather more slowly. An IMF mission arrived and took a floor of the best hotel in the capital.
That crisis has not retired. A second one has been developing closer to home.
Governments now borrow more from their own banks, pension funds and insurers. The foreign-exchange risk may shrink. The risk itself moves onto the balance sheets that hold the nation's savings. The liability has changed address. It has not left town.
And once the state becomes the most attractive borrower in the country, other borrowers notice. Usually they notice in the reply to a loan application.
The creditors have moved closer
The revised framework keeps debt-to-GDP on the panel. It stops asking it to do all the work. Gross financing needs, interest relative to revenue and domestic-debt vulnerabilities get a more explicit place in the analysis. Debt-carrying capacity is treated more carefully through growth, reserves, institutions and macroeconomic volatility.
This is less revolutionary than it sounds, and practitioners will recognise every piece of it. A finance minister already knows that a ten-year bond behaves differently from a Treasury bill maturing before Christmas. A bank treasurer knows that a government security paying a magnificent yield changes the economics of lending to a business. A pension fund knows that local-currency debt and a Eurobond are different animals, even when they share a name on the term sheet.
The framework is catching up with the plumbing. Plumbing tends to be where the flooding starts.
Same continent. Completely different debt problems.
Bubble size indicates latest available or near-term gross financing need; border/fill distinguishes financing structure. Read directionally, not as a credit ranking.
Figure 2. Nigeria is the visual oddity: a relatively modest debt stock sits beside a very large federal interest burden. Egypt's debt problem is large on both the stock and refinancing axes. South Africa carries more debt on a much deeper domestic market. Sources: IMF 2026 country reports and debt sustainability analyses; Kenya National Treasury; Seven Gates calculations. Fiscal perimeters differ. Use directionally, not as a credit ranking.
Nigeria first
Nigeria's 36.1 per cent debt ratio looks moderate. The cash burden does not. The IMF also shows claims on government rising while private-sector credit contracted in 2025. Reserves improved materially and the external position strengthened, which gives the country genuine buffers. The domestic market is therefore both a strength and a complication.
In The Grapes of Wrath, the men sent to push tenant farmers off their land explain that the bank is not a man. It is a monster that lives on profit, and the people who work for it cannot stop it eating. The modern version needs less poetry. The bank has a hurdle rate, and the sovereign has just set it.
Illustration
Picture a loan officer in Lagos with two files on her desk. The first is thick: a manufacturer in Aba wants to extend a working-capital line. It needs a site visit, a collateral valuation, a credit committee, quarterly monitoring and a quiet prayer about the exchange rate. The second file is one page long. It is a Treasury bill. It pays very well, it needs no site visit, and the borrower prints the currency it owes. She is not a villain. She is doing arithmetic. By Friday the manufacturer hears that the bank is "reviewing its sectoral appetite", which is the polite way of saying the government got there first.
At a sufficiently generous sovereign yield, the loan officer may discover a philosophical objection to factories.
That does not mean every naira lent to government is a naira denied to business. Banking systems are more complicated than that. But the sovereign yield becomes the price every private borrower must beat, and most of them cannot.
That is where debt leaves the finance ministry and walks into the wider economy.
| 36.1% Public gross debt / GDP 2025 estimate | 53.2% FGN interest / FGN revenue 2025 estimate |
Source: IMF Nigeria 2026 Article IV Consultation, Country Report 26/125.
Egypt and the calendar
Egypt reaches a similar destination from the other direction. The IMF projects public debt at 91.1 per cent of GDP in FY2025/26. More revealing, gross financing needs are expected to reach roughly 42 per cent of GDP. They are expected to stay close to 40 per cent in the near term, falling below 30 per cent by 2030 if the programme holds.
That is not merely a lot of debt. It is a lot of debt that keeps coming back.
The British once issued a bond called the Consol, a perpetual instrument with no maturity date. When the Treasury finally redeemed the last of them in 2015, some of the money being repaid traced its lineage to the South Sea Bubble and the Napoleonic Wars. A 91-day Treasury bill is the Consol's comic opposite. One can sit quietly for generations. The other insists on seeing you again before the quarter is out.
Face value tells you very little about that difference. Each Egyptian refinancing drags yesterday's borrowing into contact with today's interest rate, and today's mood. Egypt has met that mood before. When Russia invaded Ukraine in 2022, foreign investors pulled roughly $20bn out of Egyptian local debt in a matter of weeks, by the authorities' own account. Mexico learned the same lesson in 1994 with its short-dated, dollar-linked Tesobonos.
A short calendar lends you money in good weather and asks for it back in a storm.
In Hemingway's The Sun Also Rises, Mike Campbell is asked how he went bankrupt. "Two ways," he says. "Gradually and then suddenly." A heavy refinancing calendar is how gradually becomes suddenly. The calendar becomes part of the credit.
| 91.1% Public debt / GDP FY2025/26 | 42.0% Gross financing needs / GDP Near-term peak |
Source: IMF Egypt Country Report 26/224.
Ghana after the repair crew arrived
Ghana has already found out what happens when the calendar wins. Public debt fell to 48.8 per cent of GDP at end-2025, down dramatically from the post-crisis peak. The improvement reflected restructuring, fiscal adjustment, nominal growth and a stronger exchange rate.
Restructuring does not make debt evaporate. A government liability is usually somebody else's asset, and sometimes that somebody is standing outside your office. In 2023, as the domestic debt exchange took shape, retired bondholders picketed the finance ministry in Accra. Pensioners were eventually carved out of the exchange. Banks, insurers and ordinary savers carried much of the rest. Somebody always pays for a haircut, and it is rarely the barber.
After the exchange, Ghana leaned heavily on Treasury bills. In April 2026 it returned to the domestic bond market with a seven-year Treasury bond, an early step towards rebuilding term funding. This is where a bond investor should lean forward.
The interesting Ghana story is not that the debt ratio fell. Anyone with a browser can find that out. The useful question is whether the state can move from rolling short paper towards a working term curve at tolerable yields.
If it can, duration becomes valuable long before the headline ratio becomes interesting again.
Ghana's market repair
Domestic debt exchange → heavy T-bill reliance → April 2026 seven-year bond → attempt to rebuild the domestic term curve.
Source: IMF Ghana 2026 Article IV and Joint Bank-Fund DSA, Country Report 26/212.
Kenya moves the weight home
Kenya sits earlier on the same road. At end-June 2025, public and publicly guaranteed debt stood at 67.8 per cent of GDP. Domestic debt was already larger than external debt. Kenya's 2026 Medium-Term Debt Management Strategy deliberately leans on domestic financing while reducing Treasury-bill dependence and extending maturity.
Interest is where the strain shows. Kenya's budget material shows interest payments absorbing 39.8 per cent of ordinary revenue.
The obvious remedy is more revenue. Kenya has already tested the street price of that remedy. In June 2024 protesters opposing the Finance Bill broke into parliament, part of the building burned, and the president withdrew the bill. Revenue capacity is an economic variable on the spreadsheet. On the street it is a political one, and it has a breaking point.
So the debt strategy makes sense and sets a new test at the same time. Can the sovereign lengthen its funding without paying yields that keep the state expensive to finance and everyone else expensive to lend to?
That is a more useful question than whether 67.8 per cent is too high.
| 67.8% Public debt / GDP End-June 2025 | 39.8% Interest / ordinary revenue FY2025/26 |
Source: Kenya National Treasury, 2026 Budget Policy Statement and 2026 Medium-Term Debt Management Strategy.
South Africa has more road
South Africa spoils any attempt to swap one simple ratio for another. The IMF projects gross government debt in the upper seventies as a share of GDP and still rising under the baseline. Revenue is far stronger than Nigeria's, and the domestic financing system runs much deeper.
South Africa can carry more because the machinery beneath the liability is better built. Long maturities matter. Local-currency financing matters. A large institutional investor base and a deep bond market matter.
None of that makes the interest bill disappear. It means a given amount of debt behaves differently in Johannesburg than in a shallow market with a fragile currency and three months of financing visibility. The market has also shown how quickly it reprices South African politics. In December 2015 President Zuma fired his finance minister, Nhlanhla Nene, and installed a little-known backbencher. The rand and the bond market fell hard enough that the replacement lasted a weekend. Depth does not make a market docile. It makes it liquid enough to register an opinion at speed.
Market depth is part of sovereign carrying capacity.
| 77.8% Gross government debt / GDP 2025 baseline | 15.2% Average medium-term GFN Approx. % of GDP |
Source: IMF South Africa Country Report 26/034.
Zambia and the FX door
Zambia already knows what coupon day feels like when it goes wrong. In November 2020 it missed a $42.5m Eurobond coupon and became the first African sovereign to default in the pandemic. Working out the restructuring took years.
Today Zambia makes the point from the other side. Revenue excluding grants is comparatively strong and fiscal consolidation has progressed. Yet Zambia still lacks access to international capital markets and carries significant refinancing and external-debt constraints.
Non-residents held 24.2 per cent of outstanding domestic-currency government securities at end-September 2025. That runs a direct line between the local bond market and the exchange rate.
If foreign buyers stay, the government gets financing and the kwacha gets support. If they leave, both questions become more interesting at about the same time.
A high yield is not always mispricing. Sometimes the market is helpful enough to print the warning on the coupon.
Source: IMF Zambia Country Report 26/021 and Joint Bank-Fund DSA.
A small piece of arithmetic
We can formalise the argument without pretending the notation makes it more profound.
Here R is revenue capacity, g sustainable real growth, r the effective interest burden, M maturity structure, FX foreign-currency vulnerability, GFN gross financing requirement and F domestic financial-system absorption capacity.
This is an analytical identity, not an estimated structural equation. To make it useful, Seven Gates scores each component from 0 to 100 and takes the simple average.
The score is coarse on purpose. A 74 is not statistically distinguishable from a 71. A 54 is not a probability of default. The heat map matters more than the total, because two countries can reach the same number by completely different routes.
Seven Gates Sovereign Carrying Capacity Heat Map
| Country | Revenue | Growth | Interest burden | Maturity | FX resilience | GFN | Financial system | Total |
|---|---|---|---|---|---|---|---|---|
| Nigeria | 20 | 60 | 20 | 80 | 60 | 60 | 70 | 53/100 |
| Egypt | 60 | 80 | 20 | 20 | 60 | 20 | 60 | 46/100 |
| Ghana | 60 | 80 | 60 | 40 | 70 | 60 | 55 | 61/100 |
| Kenya | 50 | 80 | 40 | 60 | 55 | 40 | 55 | 54/100 |
| South Africa | 80 | 40 | 60 | 90 | 95 | 60 | 95 | 74/100 |
| Zambia | 75 | 80 | 60 | 40 | 40 | 40 | 40 | 54/100 |
Figure 3. Nigeria and Kenya arrive at almost the same total by entirely different routes. Egypt's weak point is refinancing intensity. South Africa's strength is the market beneath the debt. Source: Seven Gates Research assessment using IMF and national-authority data cited in this article. Equal weights. Scores are analytical judgements, not ratings or default probabilities.
What does 53 mean?
It does not mean Nigeria is 53 per cent safe.
Treat it as an engineering inspection rather than an exam mark. The operators at Three Mile Island had more readings than they needed. What they lacked was the habit of asking which reading might be lying. A vessel can have excellent metallurgy and a terrible relief valve. Another can have weak steel and very low operating pressure. The aggregate number forces us to walk the whole plant. The component map tells us what would actually fail.
Nigeria's weak points are revenue and interest. Its maturity profile and domestic financing capacity provide real cushioning. Kenya has better revenue mechanics but more pressure from the volume of financing the domestic market must absorb. Egypt's 46 is dragged down by the calendar. Ghana's 61 records genuine repair, with a term market still under construction. South Africa's 74 is mostly structural. Zambia's 54 pairs decent revenue and growth with weaker market access, refinancing capacity and FX resilience.
One ratio cannot do this job. Six gauges can at least start it.
Capacity is not opportunity
One variable is missing from the carrying-capacity equation.
Price.
ΔC is the change in carrying capacity, P is what the bond, currency or equity already prices, and FX is the currency effect on realised return.
South Africa has the strongest score in the group. That does not make South African bonds the best investment. The market knows a great deal about South Africa, and it charges for what it knows.
Ghana is structurally weaker. If its financing system is improving faster than investors expect, the return can be larger. Egypt has the weakest score. If gross financing needs fall quickly and maturities lengthen, it also has the most room for the market to change its mind.
A weak sovereign getting less weak can be a better trade than a strong sovereign getting slightly worse. Markets pay for surprise, and there is more surprise left in a patient than in an athlete.
Capacity tells us what can be carried. Direction tells us where repricing may begin.
Figure 4. The level tells us what the system can carry. The direction tells us where the surprise might be. Price still decides whether the surprise is investable. Source: Seven Gates Research. Direction scores are qualitative analytical judgements, not econometric estimates.
The market does not wait for the report
The new framework is not an investment. Nobody should buy a Ghanaian bond because an IMF methodology document gained several hundred pages.
The advantage lies in knowing which variables to watch before debt/GDP catches up. Rudi Dornbusch, the MIT economist who spent his career watching emerging markets break, put the timing problem best: "The crisis takes a much longer time coming than you think, and then it happens much faster than you would have thought."
Recoveries tend to run the same way. By the time either shows up in the headline ratio, the price has usually moved.
In Nigeria, change may first appear in revenue collection, domestic auction yields, bank sovereign exposure and private credit. In Egypt, the first honest signal is probably gross financing needs and average maturity. In Ghana, it is the rebuilding of the domestic term curve. In Kenya, the test is whether the long end can absorb more domestic borrowing without staying prohibitively expensive. In South Africa, the market already understands the depth of the funding base. Any surprise would come from genuine debt stabilisation or renewed fiscal deterioration. In Zambia, local-bond participation, reserves and the kwacha should be read together.
James Carville once said that if he were reincarnated he wanted to come back as the bond market, because it intimidates everybody. It also reads the data before the rest of us have finished downloading it.
By the time all of this looks beautiful in debt/GDP, somebody else may already own the trade.
Where repricing begins
Fiscal repair
Interest/revenue ↓
GFN ↓
T-bills ↓
Maturity ↑
Sovereign curve
Auction yields ↓
Term premium ↓
Duration can work
FX
Foreign demand ↑
Rollover pressure ↓
Risk premium ↓
Banks
Carry falls
Private credit becomes relatively attractive
Wider economy
Borrowing costs ↓
Credit can recover
Rate-sensitive assets
Figure 5. Sovereign repair can travel from the Treasury into the curve, then FX, banks and eventually private capital. Deterioration can make the same journey backwards. Transmission map, not a forecast.
Where the reader can make money
The people in Michael Lewis's The Big Short made their money by reading the documents everyone else had skimmed. The IMF has just published several hundred pages that most of the market will skim. That is not an investment case. It is a reading list.
The answer is not to buy the country with the best score. It is to find a financing system whose direction is improving faster than the price recognises.
Duration. If a country moves from short, expensive domestic financing towards longer maturities, lower gross financing needs and better credibility, the long end can reprice before debt/GDP falls much. Ghana is the clearest case to watch.
FX. Where non-residents own local government securities, improving sovereign credibility can support the currency through local-bond demand. Zambia makes the channel visible. The same trade holds two ways to lose money, and they tend to arrive together.
Banks. High sovereign yields can flatter bank earnings. That may reflect an excellent franchise. It may also reflect a very generous Treasury. For Nigerian, Kenyan, Ghanaian and Egyptian banks, one question deserves to become routine: how much of this profitability belongs to the bank, and how much belongs to the sovereign yield curve?
Domestic equities. If sovereign crowding-out eases, companies can reach cheaper capital again. The larger trade may then have nothing to do with the government bond. It may sit with the Aba manufacturer whose file finally gets read, and whose cost of capital falls when the Treasury stops consuming so much of the financial system.
| Country | Capacity | Direction to watch | First market evidence |
|---|---|---|---|
| Nigeria | 53 | Revenue reform versus persistent interest burden | Domestic curve, bank sovereign exposure, private credit |
| Egypt | 46 | GFN and maturity improvement | Local yields, bank exposure, FX risk premium |
| Ghana | 61 | Reconstruction of term funding | Bond-market access, T-bill reliance, cedi |
| Kenya | 54 | Domestic lengthening without permanently high long yields | Long-end yields, private credit, bank portfolios |
| South Africa | 74 | Debt stabilisation versus continued interest creep | SAGB curve, term premium, rate-sensitive equities |
| Zambia | 54 | Market access and lower FX/refinancing risk | Local bonds, non-resident participation, kwacha |
How we could be wrong
There are two easy ways to overstate this thesis.
The first is to become so taken with the new variables that we forget why debt/GDP survived. GDP remains a fair approximation of the base from which taxes must eventually come. A country cannot service debt forever by admiring its maturity profile. The indicator light at Three Mile Island was not useless. It was incomplete.
The second is to assume that better financing mechanics must produce a return. They need not. A sovereign can lengthen maturity by paying an absurd coupon. A currency can erase a perfectly good local-bond return. A bank can replace falling sovereign carry with bad corporate lending, and has done so before, in more countries than would care to be named.
So the score must move when the facts move.
| Country | What would weaken the case |
|---|---|
| Nigeria | Revenue reform stalls while interest/revenue remains structurally elevated. |
| Egypt | Gross financing needs fail to decline or maturity shortens again. |
| Ghana | Term-market access fails to deepen after restructuring. |
| Kenya | Maturity extension requires persistently punitive long yields. |
| South Africa | Debt-service costs continue rising despite the deep-market advantage. |
| Zambia | Restructuring fails to restore durable market access or FX resilience. |
Coupon day again
There will still be arguments about whether 60 per cent of GDP is too much debt, or 80, or 100. They are not foolish arguments. They are unfinished ones.
Nigeria can carry comparatively little debt against GDP and still hand more than half of federal revenue to its creditors. Egypt can be threatened by the calendar as much as by the stock. Ghana can cut its debt ratio sharply and still have to rebuild the market beneath it. Kenya can bring its borrowing home and find the domestic savings pool now carrying the weight. South Africa can tolerate a larger liability because the system beneath it is deeper. Zambia can have decent revenue and growth while market access and the kwacha remain the constraints.
The bubble chart shows where the pressure sits. The heat map shows what each system has to absorb it. The direction of travel shows what may change next. Price decides whether any of it is worth buying.
Wriston was half right. Countries don't go bankrupt. They go to coupon day, over and over, for as long as they borrow.
The bondholder still wants cash.
Research notes
The Seven Gates carrying-capacity score is an equal-weighted analytical diagnostic, not an official IMF/World Bank measure, a credit rating or a probability of default. The component profile and its direction of travel matter more than small differences in the aggregate score.
The bubble chart uses the closest available official fiscal perimeter for interest/revenue. Cross-country fiscal definitions are not perfectly identical. Where a directly comparable public gross-financing-needs series is unavailable, the chart uses the latest official near-term measure or a clearly disclosed reconstruction. It should be read as a mechanism map rather than a ranked sovereign-risk model.
Ghana, Kenya and Zambia are formal LIC-DSF cases. Nigeria, Egypt and South Africa are used as market-access comparators because the financing mechanisms motivating the reform, domestic debt, refinancing, interest burden and sovereign-bank exposure, apply beyond the LIC classification.
The Lagos loan-officer passage is a labelled illustration of a common credit decision, not a reported case. Historical references (Three Mile Island 1979, Mexico 1982 and 1994, Reinhart-Rogoff 2010 and Herndon 2013, the 2015 Consol redemption, Egypt's 2022 portfolio outflows, Ghana's 2023 domestic debt exchange, Kenya's 2024 Finance Bill protests, South Africa in December 2015, Zambia's November 2020 default) are drawn from the public record and support the argument, not the scores.
Principal sources
- IMF and World Bank, Review of the Bank-Fund Debt Sustainability Framework for Low-Income Countries: Proposed Reforms, September 2026.
- IMF, Nigeria: 2026 Article IV Consultation, Country Report No. 26/125.
- IMF, Arab Republic of Egypt: Seventh Review under the EFF, Country Report No. 26/224.
- IMF, Ghana: 2026 Article IV Consultation and Sixth ECF Review, Country Report No. 26/212.
- Kenya National Treasury, 2026 Budget Policy Statement and 2026 Medium-Term Debt Management Strategy.
- IMF, South Africa: 2025 Article IV Consultation, Country Report No. 26/034.
- IMF, Zambia: Sixth Review under the ECF, Country Report No. 26/021.
- IMF World Economic Outlook, April 2026; IMF country pages, population estimates.