YAOUNDÉ/CAIRO—Evidence of markets’ systematic overestimation of risk in developing economies has been piling up for years. The latest addition is the new release of the Global Emerging Markets (GEMs) Risk Database, which shows a sharp disconnect between developing economies’ default and recovery outcomes and prevailing market assumptions.
Investors have long associated developing economies with weak institutions, elevated political and economic risks, and frequent defaults, and have thus demanded large risk premia. When investors say that there are no “bankable” projects in developing economies, they are typically referring less to project quality than to the price of capital.
The new GEMs release—which covers more than 15,000 private-sector loans issued between 1994 and 2024 by 29 multilateral development banks (MDBs) and development finance institutions (DFIs)—upends these assumptions. It found that despite financial crises, commodity shocks, pandemics, and political upheaval, private-sector loans in emerging markets recorded default rates of 3.54%, broadly comparable to the rate for B-rated corporate bonds in advanced economies.
More important, according to the GEMs Database, when defaults occurred, lenders recovered an average of 72.9 cents on the dollar. That is much better than Moody’s global bond benchmarks and, even more so, than JPMorgan’s emerging-market recovery estimates.
The GEMs release also challenges the longstanding assumption that private borrowers in a given country cannot be less risky than the sovereign. In fact, well-structured private projects can “pierce the sovereign ceiling,” offering better credit quality than national ratings suggest. In low-income countries, sovereign ratings put default probabilities at over 31%, but in practice, private firms financed by development banks recorded a default rate of just 7.2%. In other words, while financing conditions reflected the assumption that one in three borrowers would default, the real ratio was closer to one in 14.
When risk is mispriced, and financing conditions are dictated by overly conservative assumptions, capital is misallocated, and the expectation that projects are not “bankable” becomes self-fulfilling. Elevated risk premia push required returns to levels that many productive investments cannot realistically achieve. Projects are labeled unbankable because they fail to fulfill inflated demands.
The consequences are far-reaching. As critical investments in areas like energy, transport, digital infrastructure, health care, and climate resilience are delayed—even when projects are viable—industrialization, energy transitions, and integration into emerging green and digital value chains suffer. Capital-cost differentials thus translate into development and competitiveness gaps.
The lesson is not that governance, debt sustainability, or macroeconomic stability do not matter, let alone that all developing-country investments are low-risk bets. Risk differentiation remains essential. But the aggregate evidence suggests that prevailing pricing practices often overstate risk, particularly for private-sector investments supported by DFIs. In fact, support from MDBs and DFIs materially improves outcomes, through effective project selection, supervision, governance standards, and risk mitigation. Pricing methodologies and regulatory frameworks should account for these institutions’ support.
Changing financing methods is no easy feat, as they are embedded in regulatory systems, investment mandates, and benchmark frameworks. But the costs of this inertia are high—and borne largely by developing economies. With the data now available, and the gap between assumptions and observed outcomes quantifiable, action should be taken on four fronts.
First, rating agencies should revisit the sovereign ceiling in markets where MDBs and DFIs play a significant role. A systematic review of how sovereign ratings influence private-sector credit assessments is overdue. Similarly, prudential regulators should review capital requirements for developing-country lending, recognizing the risk-mitigation benefits of MDB and DFI participation.
Third, institutional investors must update their internal risk models. As it stands, pension funds, insurers, and sovereign wealth funds often limit their exposure to developing-economy private credit. Incorporating expected-loss data that account for observed default and recovery rates could unlock substantial investment.
Finally, developing economies must continue addressing genuine sources of risk. Strengthening macroeconomic management, improving debt sustainability, enhancing contract enforcement, developing local capital markets, and creating stable regulatory environments remain essential to attracting long-term investment.
The G7, the G20, MDBs, regulators, institutional investors, and developing-country governments should treat these tasks as a systemic reform agenda, rather than a technical adjustment. This is not a request for more aid or larger concessional resources. It is a demand for capital-pricing frameworks that reflect evidence, rather than inherited assumptions.
While reforming risk assessments alone will not eliminate the global development financing gap, it could go a long way toward narrowing it, without requiring additional public resources. It would also improve the allocation of global savings and unlock productive investment at scale. Pricing capital more accurately is not simply a matter of fairness. It is a matter of efficiency, growth, and development.
Vera Songwe, a nonresident senior fellow in the Global Economy and Development Program and the Africa Growth Initiative at the Brookings Institution, is Founder and Chair of the Liquidity and Sustainability Facility. Mahmoud Mohieldin, a nonresident senior fellow in the Global Economy and Development Program at the Brookings Institution, is UN Special Envoy on Financing Sustainable Development, a former minister of investment of Egypt (2004-10), former senior vice president of the World Bank Group, and former executive director of the International Monetary Fund.
Copyright: Project Syndicate, 2026.
www.project-syndicate.org
