IFI Emergency Finance: IMF-MDB Role Reversal and Adding to the Toolkit

In the wake of the Mideast conflict, a consequential shift is occurring in the international financial architecture: MDBs are taking up the role of first responders to crises. As of early September, 80 countries have at least one arrangement under the World Bank Crisis Preparedness and Response Toolkit. Sixty-two have access to the World Bank’s Rapid Response Option (RRO), which allows countries to reprogram investment and program for results loans and use up to 10 percent of the ‌undisbursed finance for crisis response. In addition, 52 countries have Deferred Drawdown Options (DDOs), which allow them to access pre-approved lines of credit (which count against the country’s overall lending envelope).

More broadly, MDB heads have jointly pledged immediate relief for countries confronting war-related shocks in supply chains and prices. Several MDBs have announced specific financing commitments related to the conflict. The World Bank Group reports that it has made $50-60 billion available to affected developing countries, which could be scaled up to $80-100 billion if the shocks persist. The African Development Bank recently launched the Global Energy and Fertilizer Crisis Response Framework to deploy $5.1 billion in financing to offset increased energy and fertilizer prices in the continent. The Asian Development Bank has committed $4 billion in crisis financing, including $1 billion in trade finance to address price shocks. The European Bank for Reconstruction and Development has announced €5 billion in additional financing for impacted countries. The Inter-American Development Bank has expanded its total disaster risk coverage and crisis response framework to $5 billion under its Contingent Credit Facility (CCF) for natural disasters and $4.2 billion through Climate Resilience Debt Clauses (CRDCs).

The IMF has its own crisis response instruments: the concessional Rapid Credit Facility (RCF) for low-income countries and the Rapid Finance Instrument (RFI). These instruments help countries address urgent balance of payments needs caused by domestic instability, fragility, exogenous shocks, or large natural disasters. But unlike the strong uptake of the World Bank’s rapid response offerings, no countries have received these emergency finance credits from the IMF this year, with the exception of Jamaica’s RFI to support recovery from Hurricane Melissa. The direction of IMF flows is in fact shifting (as detailed in forthcoming CGD research). The 81 countries that borrowed using RCFs and RFIs during the pandemic are now beginning to repay these loans. Moreover, the IMF is returning to lower pre-pandemic access limits for these instruments.

Role reversal: What has changed?

Those familiar with traditional role assignments for the international financial institutions (IFIs) may be surprised. The IMF has historically been the first responder. Its role is to help countries in crisis return to macroeconomic and financial stability and sustainable debt positions. IMF programs combine short-term (relative to MDBs) finance with policy adjustments to help countries move back to stability. The adjustments generally target some combination of fiscal, monetary, exchange rate, financial sector, and debt policies.

The purpose of MDB finance, in contrast, is to promote long-term development objectives by supporting structural policy and institutional reforms, investments in human capital and infrastructure, and social protection. Consistent with these long-term challenges, MDB loans are much longer term: up to 35 years for IBRD loans and up to 50 years for IDA loans.

In this century, shocks with global economic and financial consequences are more common: the Global Financial Crisis, the pandemic, the Ukraine war, and now the Iran war. And they have a larger impact on poorer countries that have become more integrated into global flows of products, people, and capital. Historically, countries’ financial crises were driven principally by domestic macroeconomic policy weaknesses or regional financial contagion. The IMF was best placed to respond. Of course, there was often a need for individual countries in crisis to supplement IMF balance-of-payments support with MDB budget support. But the bulk of the finance and policy conditionality came from the IMF.

The current era confronts large swaths of countries with a whole set of additional challenges. Their own fundamentals, of course, still very much matter, but they are also grappling with frequent and large exogenous shocks not caused by their own policy decisions. Some, but not all, are climate-related. In fact, our analysis shows that non-climate-related exogenous shocks are more frequent for low-income countries (LICs) and lower-middle-income countries (LMICs): conflicts and wars with global consequences, terms-of-trade shocks (including food and fuel prices), sharp reversals in global capital flows as in the global financial crisis, and pandemics. The balance-of-payments consequences can be severe. But so are the fiscal consequences for countries already squeezed hard by debt service: total debt service on external public debt accounted for 31 percent of Africa’s total government revenue even before the Iran war.

Why is the uptake for MDB emergency finance greater than for IMF emergency finance?

The difference cannot be attributed to limitations on total IMF lending capacity—generally put at near $1 trillion. For LICs with access to concessional finance from the Poverty Reduction and Growth Trust (PRGT), the IMF Executive Board endorsed reforms in 2024 that support a long-term self-sustained annual lending envelope of SDR 2.7 billion (about $3.7 billion), more than twice the pre-pandemic capacity. Nor is it related to burdensome IMF policy requirements: the RCF and RFI come “without ex-post program-based conditionality or reviews.” There may be some differences in relative speed of disbursement. The World Bank’s prearranged finance can be executed promptly upon request by eligible countries. RCFs and RFIs must be requested by the country and approved by the IMF Board, but this can be accomplished within weeks or a month or two. Disbursement happens in a single upfront transfer.

Rather, the uptake differential may stem from countries’ limited space to add to debt stocks. IMF staff note that 33 of the 50 LICs have transitioned to regular upper credit tranche programs (IMF loans totaling more than 25 percent of a country’s quota contribution to the Fund) and may not want, or be able, to borrow more emergency finance. The World Bank’s RROs and DDOs maintain borrowing within the country’s lending limit. The RRO reallocates part of the country’s existing undisbursed balances. And the DDO activates contingent lines of credit within the country’s lending envelope.

In any case, the overall direction of change is noteworthy: the IMF is doing more regular programming and less emergency finance, and the World Bank is shifting some of its resources from regular programming to emergency finance.

Is the role reversal bad?

Country choices to move in this direction are understandable but come with real costs. The MDBs are to be commended for stepping up to a countercyclical role and helping countries offer social protection, stimulus, and recovery spending. But diverting 10 percent of undisbursed World Bank lending envelopes to crisis response inevitably means less for long-term investments in health, education, and infrastructure.

Nevertheless, evidence suggests that countercyclical support during large shocks can make a significant difference for the pace and strength of recovery. Vulnerable countries’ own fiscal resources, reserves, and policy options for dealing with frequent large shocks are severely limited, especially in the short term. The World Bank is rightly giving countries more near-term expenditure space but also more control over hard fiscal choices in times of real stress. The RRO approach gives countries a role in deciding how much long-term funding to reallocate to crisis response and how to spend it.

I would argue that the addition of the emergency finance tools from the World Bank and other MDBs is the right expansion of the overall IFI toolkit and does not undermine the fundamental division of labor between the IMF and the MDBs. We have seen the damage to long-term poverty reduction progress that follows such shocks. That damage sets back the missions of both the World Bank and the IMF.

For the IMF, a shift in the other direction can also be positive if its engagement through upper credit tranche programs actually helps vulnerable countries build firmer foundations for sustained macroeconomic stability, resilience, and growth. For LICs especially, building stability and resilience requires moving beyond crisis response to sustained support for strengthening institutions and macroeconomic policy decision-making and implementation. In that sense, long-term engagement in LICs, in and of itself, should not be viewed as inconsistent with the Fund’s mandate and should be distinguished from the problems of prolonged use of nonconcessional finance.

One can argue that the overall scale of IFI balance-of-payments and budget concessional finance is too small. That seems especially obvious for the IMF if we compare even the augmented PRGT annual finance capacity of less than $4 billion to annual IDA commitments of around $40 billion. I do not address this question here, but it is part of the focus of recently launched CGD research.

Is the toolkit complete?

We’ve seen a sharp increase in the number of countries that face severe liquidity and fiscal challenges but are not now in debt distress and do not require debt restructuring. The IMF computes debt burden indicators that assess liquidity risk vs. solvency risk. While both have risen for LICs, liquidity risk indicators have risen more sharply: in 2024–2025, 43 percent of breached debt indicators leading to high risk of debt distress were related to liquidity, compared to 14 percent in 2018–2019. A central purpose of the international financial architecture should be to help prevent those liquidity-strained countries from falling into debt default.

We need to ask whether the toolkit for such countries is fit for purpose. (This is distinct from the needs of countries for whom the only credible and adequate response is debt restructuring.) Do we have the right suite of instruments? If not, what else do we need?

One answer is that we need contract-based instruments offering temporary relief that are designed to be applicable across the broad range of large shocks. They must crowd in other major creditors—public and private—to forge a collective approach to debt relief but do so in a way that is voluntary and does not violate contracts. They must help countries avoid painful debt default. And the size of relief provided must be meaningful.

Giving countries a choice: More debt or temporary debt service relief?

The G20 Debt Service Suspension Initiative during the pandemic took a new approach to emergency relief for IDA-eligible countries—increasing net inflows by temporarily reducing debt service rather than by adding new loans. Forty-eight out of 73 eligible countries made use of it, but it foundered on two basic issues: (1) credit rating agencies viewed any temporary reductions in debt service as ex post contract changes and therefore events of default; and (2) MDBs did not participate, nor did the private sector. Major debates on comparable treatment of creditors ensued and deeply divided creditors.

Since then, MDBs have taken up this contract approach to debt relief by including “pause clauses” or debt suspension clauses (DSCs) related to climate damage, health crises, and natural disasters in their loan contracts. The World Bank, for example, has extended its Climate Resilient Debt Clauses to cover all existing loans in eligible countries (small islands and other small states) so that borrowers can defer principal and interest payments. And triggers for coverage have been expanded to include droughts, floods, and health emergencies. Some parametric triggers, viewed as arbitrary, have been dropped.

But this remains a piecemeal approach, both in terms of countries and the range of shocks covered. Nepal is too large a country to qualify, despite its devastating flood. And this approach misses some of the largest shocks—the Iran war being the most obvious current example. The Global Financial Crisis is an older example.

A welcome and pathbreaking proposal advanced by bondholders under the UK London Coalition promotes more broadly applicable triggers for DSCs in sovereign debt contracts. A CGD proposal takes this approach a step further, arguing for inclusion of standardized clauses in the contracts of MDBs, bilateral official lenders, and sovereign debt issuers—promoting both comparable treatment of creditors and coverage of a larger share of debt stocks.

Triggers for clause activation can and should be simple and objective, easily replicated across contracts, and shock-agnostic: activation should depend on the magnitude of the shock, not the exogenous source. Temporary relief should be automatic if trigger conditions in the contract are met. Relief ideally should be neutral in net present value terms. And potential relief needs to be sizable to warrant the addition of such clauses to contracts.

As one hypothetical example of relative relief size, we can look to Jamaica’s finance available for crisis response to help it recover from Hurricane Melissa in October 2025, which caused an estimated $8.8 billion in damage. Jamaica’s RFI from the IMF provided $415 million. Its catastrophe bond paid out $150 million. Diverting 10 percent of its undisbursed programmed borrowing from the World Bank as of the beginning of 2026 for financing recovery would provide about $8 million. It had pre-arranged access of up to $84 million through the World Bank's Catastrophe Deferred Drawdown Option. But $1.1 billion would have been available if its total 2026 debt service for Eurobonds and MDB debt service were suspended for a year.

For sovereign bonds, the jury is still out on whether yields would increase if DSCs were included in contracts: on the one hand, such clauses are, in effect, insurance policies for which a premium is warranted; on the other, they help countries become more resilient and avoid default, thereby reducing country risk, which should have a positive effect on yields.

Importantly, in the view of credit rating agencies, such clauses would reduce the probability of default by making countries more resilient. In its July 16, 2026 report, How debt payment pause clauses affect sovereign credit quality, Moody’s states: “We would not consider activation of a debt pause clause a default if the clause is part of the original contract.” Moody’s continues: “All things being equal, having the option to pause debt service is credit positive. The ultimate credit effect depends on the nature and severity of the shock, whether the pause clause provides significant liquidity relief and whether institutions can use them effectively as part of a credible policy response.”

Credit rating agencies do note that such clauses could have an impact on MDB liquidity positions if a large number of borrowing countries temporarily suspend payments at the same time. More analysis of these risks is needed, but our preliminary look at data for this century suggests that this would have been a significant issue only for the extreme case of the global COVID pandemic and, to a lesser extent, the global financial crisis.

Conclusion

IFIs are responding to the realities of this shock-filled era, and their roles are shifting. MDBs are devoting more lending to crisis response to lessen crisis-related setbacks to development. And the IMF is following its pandemic-era surge in emergency finance for LICs with more sustained engagement through regular programs to help them build fundamentals that make them more resilient.

MDB entry into this space is a necessary and positive development. Both the IMF and the MDBs advance their missions by reducing the consequences of major shocks for long-term growth and poverty reduction.

But MDB emergency finance is largely coming out of existing finance envelopes. And diversion of long-term development finance to crisis response is not cost-free. Moreover, other creditors must be incentivized to contribute to giving countries a breathing space and helping them avoid default.

We need another tool in the crisis response toolkit, one that frees up immediate fiscal and balance-of-payments space without requiring countries to sacrifice investments in long-term development or add to their debt stocks. Our proposal is temporary debt suspension clauses triggered by major exogenous shocks from all sources that would be included in MDB loan contracts as well as in sovereign debt issuances.

DSCs will not always be the optimal tool. Depending on debt service profiles, deferring interest for a year might just add to large debt service humps for some countries. For poor and vulnerable countries with very limited market access, more grants or highly concessional loans from the IFIs with generous grace periods would be a more effective response.

But such resources are highly scarce and must be reserved for countries without viable alternatives. For others, the option to activate temporary DSCs would be a more efficient use of resources and would enable both public and private creditors to contribute to debt relief on comparable terms.