When governments borrow from the capital markets, the credit rating agencies Standard & Poor’s, Moody’s Investors Service and Fitch Ratings (“Big Three”) issue sovereign credit ratings to guide supply and demand for the bonds. Sovereign credit ratings function like valves in the credit pipeline: the higher the rating, the lower the cost, the faster the transmission to end-users. And vice versa. Are sovereign ratings biased against economic growth in Africa?
Governments have special powers over their subjects. And credit rating agencies have special powers over the governments they rate. Sovereign ratings can impact the country’s entire supply chain of credit, to states, cities, banks, insurers, corporations, small enterprise, infrastructure, educational institutions, and consumers, with knock-on effects on labor, fiscal development and national wealth.
Creditors the world over rely on credit ratings to evaluate and price bonds—some by choice, others by regulation. One sometimes hears investors say they don’t consider ratings, but this view ignores real market dynamics. Credit ratings connect supply and demand to a market clearing price. Most investors are price-takers. The only times prices and ratings decouple are when the whole market disbelieves in the rating, or demand is thin.
Blame Africa, Or Credit Rating Models?
For several years running, the Big Three have serially downgraded African government debt. In response, a coalition of African economic and political leaders has become increasingly vocal about what they consider bias in the ratings. In September 2023, the African Union threw its weight behind a plan to launch a pan-African credit rating agency (now scheduled for mid-2025) to issue its own domestic credit risk ratings and bring a local perspective to the dialogue.
Reuters broke this news. The article said the Big Three deny bias and claim they “follow the same formula across continents.” This summer, a follow-on Reuters article succinctly summarized Africa’s predicament: “optimism has faded, washed away by a deluge of debt.”
Reuters blamed Africa. After interviewing “dozens” of rating agency employees, ex-employees and large private creditors, and reviewing “hundreds” of pages of legal documents, Reuters said they found no evidence of systemic bias—only a certain unpreparedness to rate poor countries lacking familiarity with credit rating processes.
These justifications strain credulity. The Big Three are global centenarian giants who have weathered many economic cycles, whose annual revenues match or exceed the GDP of some African countries. Neither inexperienced nor under-resourced, they simply run their businesses as they see fit.
Choice Architecture In Credit Models: Hidden Bias
It is factually incorrect to claim no documentary evidence of bias. Senior sovereign credit rating analyst declarations on the existence of bias can be searched and found publicly. Moreover, a growing academic literature that documents how Africa systematically overpays for credit access, and how ratings punish Africa while rewarding other countries for investing in capacity, merits more serious attention.
When I was interviewed by Reuters for the article, I raised choice architecture as a source of bias: the gamification of outcomes by filtering out content that could change (bias) the outcome. The journalist did not consider my points “real-world” and chose not to include them.
I shared this blog with Reuters, whose spokesperson commented: We stand by our reporting which fully met our standards for independence, accuracy and impartiality under the Reuters Trust principles.
But choice architecture is a real-world concern for credit ratings. The Big Three’s defense, that the formula for sovereign ratings is uniformly applied to all sovereign ratings everywhere, is a perfect example. Far from illustrating consistency or the absence of bias, it exemplifies the fallacy of composition, which assumes what is true for one member of a collective (rated sovereigns) is true for all members.
Moreover, it is the diametrical opposite of what I was taught as a Moody’s analyst in Asia before and during the Asian Crisis. We were told positively that culture, history, law and psychology have a role in how borrowers service their debts. We were urged, as analysts, to go beyond personal, social or cognitive biases in order to assess real repayment likelihood. I transmitted this learning in a 2011 manual I wrote for Hong Kong’s first credit rating agency analyst examinations.
Later, working as a Daubert-qualified credit ratings expert in high-profile litigations related to the Global Financial Crisis, I heard many instances where wrongdoing was denied. To reach a clear determination, each case needed to be re-analyzed from scratch to produce an independent rating judgment that could be compared to the original rating agency rating. It was—is—the only way to ascertain rating impropriety, including bias.
Sovereign Credit Ratings And Africa’s Middle-Income Trap
In the 1990s, in tandem with globalization, the Big Three expanded the orbit of sovereign ratings business from the developed to the developing world. By 2003, while Asia was recovering from its financial system crisis and the West piling into a stock-driven bubble economy, Africa was going from strength to strength on multiple indicators of real economic advancement.
S&P and the United Nations Development Programme set out to rate more African sovereigns, to enable them to tap into global private credit channels and extend their growth. The first rated African countries were colonies of France or Britain: South Africa, Tunisia, Mauritius, Egypt and Morocco. Today the total is 33, according to the UNDP.
But from the GFC through the Pandemic, African economies entered a downturn and interest costs rose. Surpluses for fiscal and social reinvestment dwindled. National finances fragilized. Waves of sovereign downgrades followed.
Too rich for the concessionary financial support of the past, African governments fell back on their limited domestic capital sources. It was a classic middle-income trap, in which credit ratings, however unintended, were a spring pin that worked against Africa’s economic growth.
A Fine Line Between Bias And Discrimination
In 2008, Ghana’s domestic long-term interest rate was 12%. In 2024, it stands at 29.85%. Nigeria’s was 9.75% in 2008, 6.25% in 2010. Now it is 27.5%. Interest rates are facts.
Between 2008 and 2024, Uganda’s sovereign rating dropped from B+ to B-/B3 (S&P/Moody’s) and Nigeria’s from BB- to B-/Caa1. Mozambique and Kenya’s ratings fell from B+ to CCC+/Caa2. Ghana and Zambia missed payments in 2022 and 2020, and their ratings dropped from B+ to Selective Default/Caa2. Credit ratings are opinions. An opinion by its very nature has bias. It reflects unconscious emotional or cognitive beliefs, or one-sided judgments. Everyone has biases. Acting on them with selectively negative consequences has a different name: discrimination.
The elephant in the dialogue about sovereign credit ratings in Africa is whether ratings by the Big Three discriminate against Africa. At the end of the day, the answer will not depend on someone’s opinion about bias in ratings. It is a question of whether sovereign credit ratings predict nonpayment risk with measurable accuracy and consistency.