In 1994, a relatively small development bank was created with just $4.3bn in capital. Within two decades, it became the world’s largest.
African multilateral financial institutions (AMFIs) can draw powerful lessons from the mandate, business model and experience of CDB. By the end of 2024 it reported $2.6 trillion in total assets and a non-performing loan (NPL) ratio of just 0.37%.
CDB weathered early turbulence
That scale and stability were neither automatic nor inevitable. CDB’s early years were turbulent. Initially, it relied heavily on fiscal appropriations and central bank refinancing. Weak repayment discipline among state-owned enterprises and an underdeveloped domestic credit system created vulnerabilities. After the Asian financial crisis in 1998, CDB’s NPL ratio soared to 43%, pushing it to the brink of insolvency.
It responded by radically restructuring. It carved out $14bn in non-performing assets, deployed debt-to-equity swaps and overhauled its operating model. It built a nationwide branch network to tighten credit control, improve project appraisal and enforce repayment discipline.
Its most consequential innovation, however, emerged at the sub-national level – provincial and municipal governments. Local governments with limited fiscal resources but strong development ambitions injected high-quality public assets into newly established platform companies with independent legal status. These entities acted as unified borrowing and repayment vehicles, accessing CDB finance for infrastructure and industrial development. Unlike many public-private partnerships, these were not designed to replace government in delivering public goods, but to structure and channel financing more effectively.
Through these vehicles, CDB pioneered “bundled” project lending – combining infrastructure projects with strong social benefits but weak standalone cashflows alongside commercially viable projects, ensuring overall loan packages remained sustainable.
This approach was first tested in 1998 in what became known as the “Wuhu model.” In Wuhu city, Anhui province, the local government capitalised a city investment company with premium assets and borrowed from CDB to finance roads, water systems and other urban infrastructure. Early repayment relied on fiscal revenues; later iterations incorporated land-use rights and land-transfer revenues as collateral.
To reinforce accountability, CDB developed internal credit rating systems for local governments and senior officials. The model was rapidly replicated nationwide, crowding in commercial banks and incubating China’s modern infrastructure financing market.
As CDB matured, it expanded beyond concessional infrastructure lending to include guarantees, equity investments and risk-sharing instruments that absorbed early-stage risk and unlocked scale.
CDB – what African banks can learn
First, African institutions today begin from a position of greater capital strength than CDB had in 1994. The Alliance of African Multilateral Financial Institutions (AAMFI) includes banks with a collective balance sheet exceeding $70bn, and the African Development Bank (AfDB) has total authorised capital approaching $320bn. Expanding asset bases to at least the CBD’s volumes over the next 30 years is entirely feasible.
Second, the development case is compelling. Average GDP per capita across Africa today is just under $3,000 – more than six times China’s in 1994. Demand for development finance from African governments and firms is at least as strong as that of Chinese local governments in the mid-1990s.
The real question is whether AMFIs and the AfDB can adapt their business models to resemble a young and agile CDB rather than traditional, risk-averse multilateral development banks.
While this deserves a much longer analysis tailored to each institution, four broad lessons stand out.
Align finance with development priorities
CDB financed China’s development plan, not isolated projects. Its lending followed priorities articulated in five-year plans, concentrating resources in sectors and regions critical for long-term growth.
For AMFIs, the equivalent is clear: treat African governments – and sub-national governments – as long-term development partners, not just borrowers. Lending portfolios should align more directly with national development strategies, industrial policies, Agenda 2063, and regional programmes such as the Programme for Infrastructure Development in Africa (PIDA), rather than remaining siloed project by project.
Be confidently cheap
Many AMFIs believe the route to concessionality is through using grants to lend just below market rates while maintaining high credit ratings, often chasing AAA status.
Yet this model can constrain leverage rather than enable it. Think of it like farming. If you plant one crop in one place, a bad season can be terminal. Plant many crops across varied fields, and a drought in one location won’t ruin the harvest. To plant many crops, CDB had to be low-cost by design – and commit to maintaining that. With high leverage, a relatively small equity base supported a large volume of assets, and scale itself reduced risk.
The lesson is counterintuitive but important: AMFIs should prioritise lower funding costs and strategic leverage rather than chasing ratings.
Be creative and less risk-averse
CDB evaluated projects through a long-term development lens rather than short-term risk perceptions. It financed projects others rejected – including those that replicated infrastructure or built capacity ahead of demand.
It built nests before the birds arrived – and in most cases, the birds came. The bank’s low NPL ratio suggests development fundamentals validated the strategy. But it required bullishness as well as creativity. During poverty alleviation programmes, for example, loans were secured against future forestry assets under the national reserve forest programme.
African multilaterals should be prepared to take informed, development-driven risk with conviction – since similar innovation is possible across Africa, particularly for cross-border projects anchored in regional strategies.
Proactively incubate projects
CDB did not wait for fully prepared projects to land on its desk. It worked upstream with local governments and firms, shaping projects to meet its standards and sharing models such as the Wuhu approach.
AMFIs can strengthen upstream engagement through technical assistance – such as the AAMFI Infrastructure Financing Facility launched in October 2025 in Luanda – and by embedding teams alongside governments to ensure projects become bankable by AMFI standards and are then financed by AMFIs themselves, rather than external actors applying different criteria.
Of course, sceptics will note key differences. CDB operates within one country, one currency and one central bank. This was not only easier administratively for the CDB, but the perceived backing of the Chinese government of CDB liabilities dramatically reduced borrowing costs. African institutions operate across multiple jurisdictions with varying governance and execution quality, and while local Chinese provincial and municipal governments too have variable quality, the fact is Beijing could intervene if necessary. The continent has no such oversight.
These differences matter – but they are not insurmountable.
AMFIs are backed by African government capital and most benefit from treaty-based preferred creditor status. This collective backing could be understood as a shared guarantee mechanism. If structured and communicated effectively, it could strengthen market confidence.
Further, progress toward greater African monetary union – and coordination on policies related to capital controls and regulatory guidance to ensure firm and household savings are channelled towards productive local use such as AMFIs – will enable more fiscal solidarity and converging oversight mechanisms, even if never as strong as China’s.
The CDB case cannot be replicated in its entirety. But it is time AMFIs fully investigate and understand its business model, in order to significantly strengthen their ability to finance Africa’s long-term transformation.
Huiyi Chen is China-Africa research and coordination analyst at Development Reimagined.
Jade Scarfe is programme manager for development finance at Development Reimagined.
Special thanks to Juliet Onyino for supporting data collection.
