When the G20 said last year that multilateral development banks (MDBs) should become ‘better, bigger and more effective’, most people assumed governments and shareholders would decide how far these banks could safely expand their lending. Instead, the biggest change came from somewhere else: a private credit rating rulebook. In October 2025, S&P Global Ratings quietly changed the way it measures the financial strength of multilateral lenders. This technical update is expected to allow MDBs to lend an extra $600–800 billion.
Many treated this as a minor technical fix. In reality, it revealed something more important: who really controls the world’s public borrowing capacity. The ability of MDBs to fund climate adaptation, pandemic preparedness, or debt relief is limited not mainly by laws or politics, but by how three private rating agencies run their models. When one of them changes its assumptions, the global system’s development capacity grows or shrinks.
How S&P’s rating methodology shapes MDB lending capacity
S&P’s revision adjusted several key parts of its model. It changed the risk weights it applies to government (sovereign) exposures under its Preferred Creditor Treatment framework. It reduced the penalties for concentrating lending in a small group of countries. And it recognised new types of contingent capital that kick in early when there is stress. Together, these changes raised the ratio that decides how much an MDB can lend without putting its AAA credit rating at risk.
S&P based these shifts on more than a decade of data supplied by the MDBs themselves. This included default data from the Global Emerging Markets database and comparisons between institutions, which showed that MDBs have an almost zero-loss history.
The revision came after the 2022 G20 Capital Adequacy Framework Review, which had identified credit rating practices as a hard limit on the international development system. That review triggered a structured process of dialogue between MDBs and rating analysts. Policy pushed MDBs to share data; that data justified changes in the models. In the end, a decade of diplomacy and reform showed up as a change in a spreadsheet formula.
S&P’s decision did not happen in a vacuum. Fitch Ratings recently updated its Supranationals Data Comparator after revising its Supranational criteria one year earlier, reflecting similar pressures to adjust its approach. Moody’s said last year that it had taken note of more transparent data on MDB callable capital and strengthened its view of shareholder support -- though it has not yet updated its criteria.
Because the credit ratings market is dominated by just a few agencies, once one of them revises its methods, competitive pressure usually pushes others to make similar changes. The scale of the impact is striking. According to the G20’s Independent Expert Group on Strengthening MDBs, years of internal reforms to MDB governance and capital structures created about $350 billion in extra lending room. A single methodological adjustment by S&P may roughly double that. This is an uncomfortable discovery: the most powerful lever in development finance sits outside the institutions that rely on it.
From credit rating reforms to a transparent global framework
The 2025 recalibration showed that this private authority can be influenced. When evidence, coordination and trust align, rating methodologies can change in response to reason. The next step is to turn this kind of engagement into something ongoing, open, and visible: a shared framework that sets out how private analytical power should interact with public policy goals.
Such a framework would not tell rating agencies what ratings to assign. Instead, it would set principles, for example: how to treat empirical-loss data, factor in shareholder backing, and account for contingent capital when assessing public financial institutions. It could also establish a forum where rating agencies, MDBs, and member governments could regularly come together to discuss how ratings are formed, ensuring that analytical independence is exercised within a shared understanding of public risk. This would bring the governance of credit closer to the rule-based systems that exist for banking supervision, accounting standards, and anti–money-laundering rules.
We already treat those systems as part of global public infrastructure; the rules that structure credit should be seen the same way. The 2025 revisions showed that the line between private analysis and public capacity is flexible and can be reshaped on purpose. Whether this becomes standard practice will depend on who writes the next methodology, and on whether governments realise that the gap between cautious balance sheet management and ambitious development goals is ultimately an issue of governance.
The reforms celebrated this year may look highly technical, but they signal a deeper shift in who holds authority in global finance. MDBs showed that, by acting together and sharing data, they can influence private rating models. Rating agencies showed that preserving their independence does not require shutting out public dialogue. The logical end point is a shared architecture in which the agencies’ analytical judgments remain their own but operate within a transparent global framework.
S&P’s October 2025 revision marks the beginning of a negotiation over who sets the limits of what is financially possible. The future of development finance will depend on whether that negotiation becomes a permanent feature of the system, and on whether the governance of credit moves from ad hoc discretion to deliberate design.
