The Fund Returns. Here Is What It Asks For.

Section III · The Ballot and the Bill · The World in 2027

The Fund Returns. Here Is What It Asks For.

On 12 November the International Monetary Fund set out three things Senegal must do before a 2.2 billion dollar programme reaches its Board. Parliament has refused one of them. Across the Fund's caseload the terms are published, the arithmetic is public, and the argument is about what compliance actually buys.

A government finance ministry negotiating room with officials seated around a table, The Fund Returns Here Is What It Asks For, The Meridian

On Thursday 12 November, the International Monetary Fund's communications director, Julie Kozack, listed what Senegal must do before its new programme reaches the Executive Board. Implement the policies agreed at staff level, including completion of a budget framework. Take corrective measures for the misreporting of hidden debt. Secure financing assurances from its partners. The staff-level agreement, worth 2.2 billion dollars over thirty-six months, was announced on 2 September. Ten weeks later, no money has moved.

Senegal's National Assembly has rejected an amended 2026 budget. A parliamentary committee has asked the finance ministry to withdraw the 2027 budget bill. Both are central to the programme. Asked whether the president might adopt a budget without parliament, Kozack declined to comment.

That exchange is the whole subject in miniature. A Fund programme is not a loan with a covenant attached. It is a sequence of dated commitments, most of them legislative, each tested before the next tranche is released. Governments sign them because the alternative is worse. Then they have to pass them.

This article sets out what the Fund is actually doing going into 2027: who owes it money, what that money costs, what the conditions say, and what the evidence, including the Fund's own, says the conditions achieve.

The Caseload Is One Country and Everybody Else

At the end of September 2026, the Fund's five largest debtors owed, in order, Argentina, Ukraine, Pakistan, Egypt and Ecuador. The distance between first and second is the fact that matters.

Chart 1 · Who Owes the Fund
Argentina owes the IMF roughly four times as much as the next largest borrower.
Argentina56.8
Ukraine14.4
Pakistan10.8
Egypt10.6
Ecuador9.4
IMF credit outstanding, billions of US dollars, as at 30 September 2026. Source: IMF credit outstanding data, compiled by Visual Capitalist. Argentina's exposure is equivalent to 8.2 per cent of its GDP, Ukraine's to 6.4 per cent and Ecuador's to 6.8 per cent.

No African sovereign appears on that list. Africa's relationship with the Fund is different in kind: many programmes, each small, most of them concessional, and concentrated in countries with no alternative lender. In the financial year to 30 April 2026 the Fund approved roughly 29 billion in Special Drawing Rights of new requests, about 39.8 billion dollars. Five of those approvals were under the Resilience and Sustainability Facility, worth about 1.8 billion in SDRs between Burkina Faso, The Gambia, Jordan, Liberia and Pakistan.

What It Costs

The price of Fund money is not a single number. It is a stack, and each layer is published.

The base is the SDR interest rate, which moves weekly with short-term yields in the five basket currencies. For the week of 9 to 15 November 2026 it is 3.119 per cent. On top of that sits a margin of 60 basis points, cut from 100 in the reform that took effect on 1 November 2024. Borrowers holding credit above 300 per cent of their quota pay a further 200 basis points. Those still above that threshold after thirty-six months, or fifty-one under an Extended Fund Facility, pay another 75.

Chart 2 · The Stack
A heavily exposed borrower pays 6.47 per cent. The poorest members pay nothing. Congo pays 9.5 on the market.
Basic rate of charge3.72
Above 300% of quota5.72
Above 300%, past 51 months6.47
Concessional, poorest0.00
DRC Eurobond, 10-year9.50
SDR interest rate, 3.119% Margin, 60bp Level surcharge, 200bp Time surcharge, 75bp Market
Annual rates, per cent. SDR interest rate for the week of 9 to 15 November 2026; charges and surcharges per the IMF's published policy, effective 1 November 2024. A service charge of 50 basis points applies to each drawing, and commitment fees of 15 to 60 basis points apply to undrawn balances; neither is shown. The Democratic Republic of the Congo's debut ten-year Eurobond priced at 9.5 per cent in April 2026. Sources: IMF Finance Department; IMF factsheet on lending; African Business.

Two readings of that chart are both correct. The first is that the Fund is cheap. A sovereign paying 6.47 per cent on its most expensive Fund credit is paying three percentage points less than the Congo paid international investors in April, and the Congo's issue was four times oversubscribed, which was treated as a success.

The second is that the Congo is doing both at once. It has had an Extended Credit Facility and a Resilience and Sustainability Facility since January 2025, worth about 1.73 billion dollars and 1.04 billion respectively, drawn on concessional terms. In April it borrowed 1.25 billion dollars from the market at 8.75 and 9.5 per cent. Concessional finance did not displace expensive finance. It sat alongside it.

The Fund is the cheapest money a distressed sovereign can get. It is also the money that comes with a legislature attached.

The Surcharge Is Mostly One Country

Surcharges were introduced in 1997 to discourage large and prolonged use of Fund resources. They have become, in practice, a transfer from a very small number of borrowers.

Chart 3 · Where Surcharges Come From
Argentina has paid 54 per cent of all IMF surcharges since 2009. Five countries account for 95 per cent.
54% ARGENTINA
Argentina 54% Ukraine 16% Egypt 14% Ecuador 6% Pakistan 5% All others 5%

Argentina has paid about 5.0 billion in Special Drawing Rights in surcharges since 2009, roughly 6.85 billion dollars. Surcharges are 42 per cent of everything it has paid the Fund in charges, and it is projected to keep paying them until 2031.

Shares of total IMF surcharges paid, 2009 to 2026. Argentina's 54 per cent share and the five-country total of 95 per cent are as reported; the remaining shares are derived from the same SDR figures and rounded. Source: Francisco Cantamutto for the Center for Economic and Social Rights, October 2026. CESR campaigns for the abolition of surcharges, and this is its own calculation from IMF data.

The October 2024 reform was real. The margin fell from 100 basis points to 60. The level-based threshold rose from 187.5 per cent of quota to 300. The time-based rate fell from 100 basis points to 75. Commitment fee thresholds moved up. The Fund estimated the package would save borrowers about 880 million in SDRs a year, roughly 1.2 billion dollars, and cut the number of surcharge-paying countries in the 2026 financial year from twenty to thirteen.

It did not abolish surcharges, and the countries that were paying the most are still paying. By mid-2026, on CESR's count, twelve countries were still being surcharged, and borrowers were due to pay about 1.68 billion dollars in surcharges across the year. The Fund's position is that surcharges build precautionary balances and discipline large exposures. The counter-position is that they fall hardest on countries already in difficulty. Both are arguments about the same arithmetic.

What the Conditions Actually Say

Conditionality is not a tone. It is a table, published in every staff report, divided into criteria whose breach stops money and targets whose breach does not.

Chart 4 · Binding and Not Binding
The monetary and fiscal numbers are performance criteria. The social spending floor usually is not.
 
Performance criterion
Indicative target
If missed
Disbursement stops unless the Board grants a waiver
Recorded in the staff report; the review proceeds
Central bank
Zero ceiling on direct advances to government, continuous
—
Reserves
Floor on net international reserves, tested each quarter
—
Budget
Floor on the primary or domestic fiscal balance
—
Social
—
Floor on pro-poor and social protection spending
Generalised from the conditionality tables of the Ethiopian, Ghanaian and Congolese arrangements. Sources: IMF country reports and Executive Board press releases, 2023 to 2026.

Ethiopia is the clearest case. Before its 48-month Extended Credit Facility of 2.556 billion in SDRs could go to the Board, the authorities had to float the birr. They did so on 29 July 2024, the day the Board met. The currency fell from 57.48 to the dollar to 74.73 within days. A zero ceiling on direct advances from the National Bank of Ethiopia to the federal government became a continuous performance criterion.

Ghana's arrangement carried the same zero ceiling on central bank credit to government, a floor on net international reserves and quarterly indexation of electricity and water tariffs.

Compliance and Outcome Are Different Things

Ethiopia has complied. Its fifth review was completed on 1 July 2026, releasing about 464 million dollars and taking cumulative disbursements to roughly 2.65 billion. The programme is on schedule.

By 3 September 2026 the birr was 161.05 to the dollar. Independent bureaux were quoting near 177 in August. A dollar now costs about 2.8 times what it cost on the morning of the float. Headline inflation, which had fallen to 9.7 per cent in December 2025, was 15.3 per cent in July 2026.

None of that is a programme failure in the technical sense. The reviews pass because the criteria were met. It is a reminder that the criteria measure policy actions, not the welfare of the people living through them.

The Fund's own evaluator has said as much. The Independent Evaluation Office, reviewing arrangements from 2008 to 2019, found that growth in programmes consistently fell short of programme projections, and that the shortfall was larger in non-concessional programmes than concessional ones. About half of programmes undershot their own initial growth projection by half a percentage point or more. About a quarter undershot by more than one and a half points. The IEO attributed much of this to modelling error, particularly in assumptions about fiscal multipliers.

The academic evidence is split and should be reported as split. Stubbs, Kentikelenis, Stuckler, McKee and King, examining sixteen West African countries and Fund programme documents from 1995 to 2014, found that each additional binding policy reform reduced government health expenditure per capita by 0.248 per cent, with a confidence interval from 0.435 to 0.060. Their mechanism is public sector wage bill ceilings and hiring restraint reaching health ministries.

Against that, Balima and Sy, using entropy balancing across 106 developing countries from 1970 to 2016, found that Fund programmes significantly reduce the likelihood of subsequent sovereign default, and that countries entering programmes tend to see modest improvement in credit ratings and falls in debt and deficits relative to output.

Both can hold. A programme can lower the probability of default and compress health spending in the same five years. The disagreement is about which of those is the thing being measured.

The Concessional Window Is Too Small

One figure in the Fund's own annual report deserves more attention than it has received. The Poverty Reduction and Growth Trust, which is where almost all low-income borrowing happens, expects demand of 3 to 5 billion in SDRs a year in 2026 and 2027. Its estimated lending capacity is 2.5 to 3.0 billion.

Chart 5 · The Gap
Expected demand for concessional lending exceeds capacity in both 2026 and 2027.
012345
Expected demand3.0 – 5.0SDR bn
per year
Lending capacity2.5 – 3.0
SDR bn
Poverty Reduction and Growth Trust, annual figures in billions of Special Drawing Rights for 2026 and 2027. Undisbursed PRGT lending commitments stood at about 27 billion in SDRs at 30 April 2026. Source: IMF Annual Report 2026.

At the top of the demand range the shortfall is 2 billion in SDRs a year. That is not an accounting detail. It is the mechanism by which a low-income country that needs concessional money in 2027 does not get it, or gets less of it, and makes up the difference on terms like the Congo's.

The terms of that window have also changed, and the change is not widely understood. Since 1 May 2025 the PRGT has charged a tiered rate on new arrangements. The poorest eligible members, about half of the group, still pay zero. The rest now pay a positive rate, concessional but no longer free. Credit already outstanding stays at zero.

Who Decides

Quota determines voting power, access limits and the cost thresholds in Chart 2. It has not been realigned in over a decade.

Chart 6 · Quota Shares
Fifty-four African member states hold a combined quota share smaller than China's.
United States17.4%
China6.4%
All African Union members5.2%
IMF quota shares, per cent of total. African Union member states hold 5.2 per cent of quota and 5.7 per cent of votes; the vote share is higher than the quota share because basic votes are distributed equally among members. Source: Development Reimagined, Reforming the IMF Quota System and World Bank Voting Shares, June 2025.

The Sixteenth General Review of Quotas was approved by the Board of Governors on 15 December 2023. It raises every member's quota by 50 per cent, leaving relative shares untouched, and would take total quotas from 476 billion in SDRs to 715.7 billion. It is not in force. It requires written consent from members holding at least 85 per cent of quota, which means it requires the United States, and the United States has not consented.

The Executive Board has extended the consent deadline repeatedly. On 8 May 2026 it extended it again, by six months, to 15 November 2026. That is three days after this article's reporting date.

Africa did gain something in the same cycle. On 1 November 2024 the Executive Board expanded from twenty-four chairs to twenty-five, creating a third sub-Saharan African constituency, taken up by Côte d'Ivoire's Wautabouna Ouattara. A twenty-fifth seat at a table where voting shares were unchanged is a real improvement in workload and a modest one in power.

What 2027 Holds

Senegal. The hardest case. Hidden debt that analysts put near 13 billion dollars, discovered in 2024; debt at 132 per cent of GDP at the end of that year; bonds that traded below fifty cents. A staff-level agreement exists. Board approval requires a budget that parliament has so far refused to pass, corrective action on misreporting, and financing assurances from creditors whose official committee has not yet been formed. Each is a separate point of failure.

Kenya. No programme since the previous arrangement lapsed in April 2025. Reporting in October 2026 indicated a new deal is unlikely before the general election on 9 August 2027, roughly eleven months out at the time, with the Fund reluctant to commit so close to a vote and Nairobi turning to the World Bank for concessional money instead. A country of 56 million is running its 2027 fiscal year without Fund support, by choice on both sides.

Ghana. On 5 November 2026 the Board completed the sixth and final review of Ghana's 39-month Extended Credit Facility, releasing 265.9 million in SDRs, about 371 million dollars. The arrangement is finished. Ghana now moves to a 36-month Policy Coordination Instrument, which lends nothing and monitors everything. The Board's accompanying conditions include holding a primary surplus of 1.5 per cent of GDP in 2026, recapitalising the Bank of Ghana by 2032, and continuing good-faith negotiation with remaining commercial creditors.

Ukraine. The 2023 Extended Fund Facility no longer exists. On 26 February 2026 the Board cancelled and replaced it with a new 48-month arrangement of SDR 5.9353 billion, about 8.1 billion dollars and 295 per cent of quota, which anchors a 136.5 billion dollar international financing package. It runs to February 2030, with reviews falling due through 2027. At 14.4 billion dollars it remains the second-largest exposure on the Fund's books, and the Group of Creditors has committed to a definitive debt treatment before the arrangement's final reviews.

Argentina. The 48-month facility approved in April 2025 runs to 2029, with reviews through 2027. More than half of all surcharge revenue depends on it.

What We Conclude

The Fund is the cheapest lender a distressed sovereign has, by a margin of roughly three percentage points against the market, and its concessional window is cheaper still. That is the strongest thing that can be said for it, and it is not a small thing.

Three qualifications follow. The window is undersized against its own projection of demand for 2027, by as much as 2 billion in SDRs a year. The governance that sets access and cost has been frozen for a decade by a consent threshold one member can hold, and the deadline for releasing it falls this week. And the conditions attached to the money are enforced asymmetrically: the monetary and fiscal numbers are performance criteria, the social spending floor usually is not, and the Fund's own evaluator has documented that programme growth routinely undershoots the projection on which the adjustment was calibrated.

The countries signing these agreements know all of this. They sign anyway, because the alternative is 9.5 per cent, or nothing.

What to Watch

First, whether the 15 November consent deadline passes without the Sixteenth Review entering into force, and whether the Board extends it for a further six months. A fourth extension would confirm that quota reform is now a formality nobody expects to complete.

Second, whether Senegal's budget reaches the statute book. If the National Assembly holds out, the programme does not go to the Board, and a sovereign at 132 per cent of GDP enters 2027 without it.

Third, whether Kenya's abstention holds through the August election. A government that gets through a general election without a Fund programme has demonstrated something about the necessity of Fund programmes, whatever the fiscal cost.

Fourth, whether the PRGT's financing gap is closed or absorbed. If it is absorbed, it will be absorbed by rationing, and rationing in a concessional window is invisible until a country that needed money turns up in the Eurobond market instead.

A Fund programme is a transaction in which a government exchanges a measure of control over its budget for money it cannot raise elsewhere. That exchange is sometimes the best available option. It is never a neutral one, and the record of what it delivers is better documented by the Fund's own evaluator than by either its defenders or its critics.

Data notes: Senegal's requirements for Board approval are from remarks by IMF communications director Julie Kozack reported on 12 November 2026; the 2.2 billion dollar staff-level agreement was announced on 2 September 2026, and the debt figures, including 132 per cent of GDP at end-2024 and the roughly 13 billion dollar analyst estimate of hidden debt, are from contemporaneous Reuters reporting. Credit outstanding at 30 September 2026 is from IMF data compiled by Visual Capitalist; the IMF's own weekly ledger could not be read directly, so the figures are reported at one remove and the date is stated. The SDR interest rate of 3.119 per cent is the IMF's published rate for the week of 9 to 15 November 2026. Charges, surcharges, commitment fees and service charges are from the IMF's lending factsheet and from Press Release 24/385 of 21 October 2024, which records the Executive Board's conclusion of the review on 11 October 2024 and the effective date of 1 November 2024; the research supplied to this desk cited Press Release 24/365 for the same decision, which is not the correct number, and the figure has been taken from the IMF's own page instead. Surcharge shares since 2009 are from Francisco Cantamutto for the Center for Economic and Social Rights, October 2026, an organisation that campaigns for their abolition; the Argentina share of 54 per cent and the top-five share of 95 per cent are as published, and the intermediate shares are derived from the same SDR figures. Congolese programme and Eurobond figures are from the IMF press release of 26 June 2026 and from African Business, April 2026. Ethiopian exchange rate, inflation and disbursement figures are from Rio Times reporting of 4 September 2026 drawing on National Bank of Ethiopia and IMF data. Ghanaian figures are from the IMF statement of 5 November 2026. PRGT demand, capacity and undisbursed commitments, total quotas, the Sixteenth Review status and the 15 November 2026 consent deadline are from the IMF Annual Report 2026 and from the IMF policy paper of 13 May 2026 recording the extension approved on 8 May 2026. Quota shares are from Development Reimagined, June 2025. The Independent Evaluation Office findings are from its 2021 evaluation of growth and adjustment in IMF-supported programmes. The health expenditure estimate is from Stubbs and others, Social Science and Medicine, volume 174, 2017. The default-probability finding is from Balima and Sy, IMF Working Paper 2019/002; that paper reports statistical significance without publishing an effect size, and none is asserted here. The research supplied to this desk also stated that concessional PRGT lending remains at zero interest for all eligible members; that is no longer accurate, and the tiered rate introduced on 1 May 2025 is reported instead. Claims in that research about which Congolese indicative targets were missed at end-December 2025 could not be confirmed against the Executive Board's own statement and have been omitted. Correction, 13 November 2026: an earlier version of this article stated that Ukraine's Extended Fund Facility was approved in March 2023 and runs to March 2027, with final reviews in the first quarter. That arrangement was cancelled and replaced by the Executive Board on 26 February 2026 with a new 48-month arrangement of SDR 5.9353 billion running to February 2030. The entry has been corrected.

Content Protected

This article is the original work of The Meridian and is protected by copyright. Please link to it rather than reproducing it.