> ## Content Index
> Fetch the complete content index at: https://dennismatanda.ghost.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# My hopes for the Africa Credit Rating Agency, AfCRA
- URL: https://dennismatanda.ghost.io/my-hopes-for-the-africa-credit-rating-agency-afcra/
- Published: 2026-10-12T19:57:00.000Z
- Updated: 2026-10-24T15:25:26.000Z
- Description: Africa’s pension funds, insurers and banks hold about US$4 trillion, by the APRM’s estimate, and less than 5 per cent of it is in anything rated. Foreign-currency bonds cannot feed another agency on fees. The neglected work is local-currency debt, where a factory actually borrows.
- Author: Dennis Matanda
- Tags: Essays

On 7 October, in Port Louis, the African Union launched the Africa Credit Rating Agency, AfCRA. It is privately owned and designed to address an old grievance: Africa pays too much for the money it needs. We hope an African agency, using African knowledge, adds to African agency. But the prices it wants to change are set by computers at pension funds in Copenhagen and Connecticut, and by the regulator’s rule book in Accra and Kampala. The chairman of the African Union Commission said as much at the launch: the agency alone will not bring Africa’s cost of capital down. It can change the list of names those computers read.

I should say where I stand. I have always wanted a better price for the capital that feeds Africa before an agency could argue for it. And by better, I always meant true. Before I became an advisor at a development finance institution, I spent significant time advocating for better everything in Africa. Those are stakes, and the reader should weigh them. They are also why I write: this is the bar I expect any institution I serve to meet.

The Union asked AfCRA for transparency and accountability in African markets, better market intelligence, better-informed investment decisions, and a stronger African voice. The African Peer Review Mechanism, which designed it, added the measures: rate the unrated, deepen domestic markets, use an Africa-oriented scale. Those are the agency’s words. We hold it to them. We are not asking for a kinder mark. We are asking the market to retire the lazy eye that saw one risk in three African countries hundreds of miles apart.

**Domestic markets, and the unrated**

Africa’s pension funds, insurers, and banks hold about US$4 trillion, by the APRM’s estimate, and less than 5 per cent of it is rated. Foreign-currency bonds cannot feed another agency on fees; Moody’s owns the largest local agency, GCR, and S&P is buying most of Nigeria’s Agusto. The neglected work is local-currency debt, where factories actually borrow. As Misheck Mutize puts it, the agency’s main job is to steer African savings into African assets.

Then there are the twenty-three countries with no rating at all. The brief has already chosen: the issuer pays. An agency that waits to be hired will rate them when they ask, and most will ask late, because a rating needs current accounts, every debt and guarantee listed, and a ministry that answers questions. The agency should say now whether it will rate a government uninvited, and that most first ratings will be low. That is no failure. A low rating that can be checked is a place on the list. No rating is not.

**The scale**

Every buyer will ask whether AfCRA uses the same yardstick as everybody else. The brief’s added tests, budget-process transparency and data transparency, are what the market wants. Its own meaning for the letters is not. If its B means what a B means in New York, a second opinion adds information. If its B means something else, the agency must print the conversion table, and the default record behind it, or the lower rating will set the price. The agency must be legible to the terminal in Copenhagen and the pension trustee in Lagos.

**Market intelligence**

Development banks have pooled their loan files since 2009\. Private borrowers in sub-Saharan Africa defaulted at 5.5 per cent a year, against 3.3 across the emerging world, and the banks still recovered 72 cents on the dollar, 95 from governments. That record beats the rumour, and the agency should say so. But a careful development bank is not the market: its loans had covenants, and governments paid it first. The agency’s job is to give the market a number it can lose money on and still defend to its board. That is market intelligence.

**Independence**

AfCRA will also rate Africa’s development banks, and the one I serve, the Trade and Development Bank Group, is AAA in Beijing, A-minus in Tokyo, BBB-plus at GCR, and below investment grade at Moody’s and Fitch. Afreximbank was cut by Fitch in June 2025, ended the relationship, and was cut again as Fitch withdrew, over whether its Ghana loan enjoyed preferred-creditor status, which the bank and the APRM dispute. Same banks, same accounts, different letters.

How should AfCRA mark them? Not by loyalty, and not by copying. A development bank’s rating rests on its own capital, the strength of its owners, and whether borrowers really do pay it first. The first is in the accounts. The second is a hard fact: many owners are rated single-B or lower. The third is a record. Which member states kept paying these banks in full while others were in default? Publish that table, loan by loan, and the quarrel with Fitch is settled by evidence, whichever way it falls. The Alliance banks should publish it themselves, without being asked.

One conflict is built in. Governments may not own shares, but the brief lets “designated institutions” and multilateral banks hold them, and those banks are owned by the governments to be rated. Any bank with a stake should sit outside the rating committee and be named on the rating. If AfCRA then marks such a bank lower than it would like, it has earned the right to be believed about everyone else.

**Transparency and accountability**

Publish a method an outsider can reproduce. Record the committee’s votes and have the method audited. Where data is missing, mark the gap, not the silence. Publish, every year, whether last year’s ratings were right. Then mark a powerful government, or a powerful bank, down when the file says so. Independence, says Daniel Cash, who has advised the project, “is demonstrated at the point of disagreement”. Until the first bad rating a member state does not like, the agency has a promise, not a record.

**Voice**

The last objective is the one we would warn the agency’s friends about. The UN’s Special Adviser on Africa told the launch that African capital must use its ratings: “We cannot ask global capital to trust what African capital treats as optional.” That is the one way to kill the thing. An agency whose customers are told to use it is paid whether it is right or not, and money paid regardless stops looking. That is the sixty-year habit this continent is trying to break. Its first users should come because its hit rate is published and good, and its first recognition should come from a pension regulator, based on that record. The computer reads only what its regulator recognises; a voice is earned the same way.

The market is not refusing Africa out of contempt. Contempt at least looks. This is the shrug of a buyer given nothing to check. Africa does not need an agency that tells a better story. It needs one that makes the real balance sheet readable and stands by the number when it's ugly. A true price is not always a lower price. AfCRA can deliver transparency and intelligence by itself. Better decisions belong to the buyers. The voice has to be earned. The computer did not come to Port Louis. It will read what this agency publishes, in a form it can use, or it will not read the agency at all. We will be reading too, and we have said where we sit.