Is there an "Africa penalty" in sovereign ratings by credit agencies?

On the information problems associated with risk analysis in Africa

Thank you for being a regular reader of An Africanist Perspective. If you haven’t done so yet, please hit subscribe in order to receive timely updates.

I: There are signs that creditors are willing to bet on African economies again

African economies have had a decade of low growth and stagnant per capita incomes. However, as I keep arguing, this time is different. The current regional economic slowdown is unlikely to precipitate the sort of disastrous outcomes witnessed in the 1980s and early 1990s.

Moving forward, it will be important for African governments to avoid total fiscal meltdowns and associated austerity-obsessed structural adjustment programs that characterized the 1980s. The imperative during this difficult but transitory period ought to be to protect the human development gains in healthcare, education, and nutrition made over the last 30 years. For the most part, African economies are in a position to grow out of their current fiscal squeeze.

That is why Benin’s and Cote d’Ivoire’s recent oversubscribed Eurobond is a good sign that credit-worthy-ish (“junk” status) African states’ fiscal squeeze might be coming to an end. Further encouraging signs came on February 7, 2024 when Kenya’s notice that it intends to go on the market saw its yields plummet from 16% to under 10%. Nairobi’s new Eurobond issue will be priced on February 12, 2024.

Abidjan, which currently has a rating of BB- (Fitch), had gone to the market for $2.6b and got interest from over 400 investors with offers exceeding $8b — “the highest order level ever recorded by a sovereign in West Africa.” According to Bloomberg, the Ivorians sold “$1.1 billion via a sustainable bond maturing in 2033 and $1.5 billion in conventional bonds due 2037.” For its debut dollar bond, Benin was on the market for $750m and got offers to the tune of $5b.

Cote d’Ivoire, in particular, is emerging as the star West African economy as the region continues its post-pandemic recovery (especially in light of Senegal’s worsening political crisis). With growth projected to hit 6.6% this year and average above 6% over the next five years, the country has enough runway to service its debts comfortably into the medium term. The only thing that could get in the way, of course, is the Ouattara succession in October 2025.

Trends in yearly debt service obligations (interest plus principal payments on public and publicly guaranteed debt) to external lenders across Africa (figures based on debt obligations as of December 2022). Source: ONE.ORG

Looking at the trend data, the shocks associated with COVID and geopolitics could not have come at a worse time for several African economies. Debt service obligations in the region were scheduled to rise sharply and peak just when high interest rates and perceived credit risk made it impossible to rollover government debt. There is good reason to believe that the region’s economies would’ve done better at servicing their debts had it not been for the pandemic and other shocks experienced over the last three years.

Eurobond maturities over the next 20 years. Source: Gregory Smith

Over the last two decades several African countries accumulated private debt (largely Eurobonds) due to the speed of disbursement and lack of policy prescriptions. Such debt rose to about 38% of the total capital flows into African economies. The ability to borrow created fiscal space for African governments to invest in ambitious projects that matched their domestic priorities and political economy constraints. It is also true that in some cases low borrowing costs fueled insane levels of corruption (see here and here). It is those debts that have come to haunt a number of countries as currency depreciations, rising interest rates, declining commodity exports, and domestic policy mistakes conspired to make it impossible to rollover debt.

A breakdown of private, bilateral, and multilateral external creditors to African sovereigns. Source: ONE.ORG

There is no doubt that many African states could have done a better job with regard to debt management in the heyday of high commodity prices — from prudently investing borrowed money to properly spreading out repayment schedules and accounting for forex movements and interest rate risk. However, it’s also true that the inflationary and interest rate pressures caused by high-income countries’ responses to COVID and other shocks made things worse just as debt servicing obligations were about to peak in the region.

When Ghana defaulted in December 2022, it was scheduled to witness a 54.2% increase in externals debt service obligations the next year (and had averaged $2.4b in external debt servicing over the preceding 5 years). Ethiopia defaulted in December 2023, a year that saw a 53% increase in debt servicing obligations, and an even bigger increase slated for 2024. Zambia’s default in late 2020 was driven by both COVID-related liquidity constraints and the fact that debt service obligations were scheduled to balloon by over 368% in 2023.

In 2024 the most important economy to watch will be Kenya. It’s debt service payments in this cycle will peak with a $2b Eurobond payment due in June. As noted above, Nairobi looks set to issue a new Eurobond in the first quarter of this year. A number of conflict-affected states across the Sahel might also go through technical defaults due to sanctions (but those will be states of exception).

Overall, it is important to note that the underlying drivers of debt distress or default risk in the region over the last three years have been related to liquidity constraints rather than generalized insolvency. The liquidity constraints, in turn, have been the result of domestic policy mistakes and political economies, as well as (and perhaps mostly because of) external factors beyond the control of African governments.

II: On the “Africa penalty” in sovereign credit ratings

Over the last three years perceived negative bias towards African sovereigns by CRAs has generated critical analyses from researchers and several organizations including the United Nations, the International Monetary Fund, and the African Development Bank (see here, here, here, here, and here).

Beginning with South Africa in 1994, the number of African states rated by the sovereign credit rating agencies (CRAs) has increased to 32. Having formal ratings by the top three agencies — Moody's, Standard & Poor's, and Fitch — enabled African sovereigns and some firms to raise money on global credit markets. However, it also exposed them to the signaling power of the CRAs and related impacts on their cost of borrowing.

Source: Misheck Mutize and McBride Nkhalamba (2021)

The criticisms of credit rating agencies can be summarized into for broad categories:

• Ideological opposition to government spending leads them to punish developmentalist fiscal policies that, in actual fact, bolster African governments’ credit worthiness.

• The CRAs scarcely invest in knowing rated African economies despite collecting fees from regional governments (lack of in-country staff, in particular, limits their ability to provide “objective” qualitative input into their rating decisions).

• A lack of sufficient knowledge of African economies mean that rating agencies have a built-in downward bias against African sovereigns. Downgrades tend to be pro-cyclical — especially during global crises — in ways that are self-fulfilling; while African sovereigns seldom get positive rating upgrades when global conditions improve. Furthermore, downgrades of individual countries tends to have regional effects on borrowing costs. It’s almost as if both CRAs and creditors think that Africa is a country.

• The high baseline borrowing costs associated with the structural biases against African sovereigns unnecessarily raise their financing costs, a fact that endogenously raises their default risk.

These charges have merit. As Fofack observes:

Until steps are taken to transcend historical biases and the one-size-fits-all approach to risk assessment, as well as to integrate Africa’s brightening realities and diversity of circumstances (growth prospects, stage of development and economic complexity) into sovereign risk models, the region—in spite of accounting for the lowest share of global sovereign debt—will, more often than not, be on the verge of debt distress because it is paying too much interest on its relatively low and even marginal stock of external debt…

The combined external public debt stock of Africa, which accounts for around 17 percent of the world’s population, is less than 1.5 percent of the total global sovereign debt, which was about $73 trillion at end-2019 and will be significantly higher after the extension of disproportionately larger discretionary fiscal support by advanced economies...

Understandably, the CRAs have pushed back hard against these charges. For example, Moritz Kramer, formerly of S&P, argues that sovereign rating agencies are actually biased in favor of African countries. As evidence, he notes that:

Sub-Saharan African sovereigns rated in the B category between 2010 and 2023 defaulted in 22 per cent of all cases within five years. The respective global number stands at a long-term average of only 16 per cent. Over at Moody’s, the observed default ratios look similar at 30 per cent for sub-Saharan African sovereigns and 15 per cent for its global average.

Hanah Ryder, of Development Reimagined, disagrees. She argues that observed rates of default in African states are endogenous to the sovereign ratings. In other words:

downgrades - just by their announcement - lead to defaults (ie the “self fulfilling prophecy” nature of credit ratings/ asymmetric information problem)

Ryder has a point. Consider the case of Ghana. It is certainly true that the current administration mismanaged public finances by over-leveraging the economy. However, it is also true that the downward cascade of the country’s fiscal position was, in part, a reaction to credit downgrades caused by external shocks unrelated to objective measures of sovereign risk before the effects of the shocks set in:

Before the pandemic, the country’s heavy borrowing included $13.2 billion in Eurobonds – representing 17% of its GDP. International investors piled in, with a 2020 Eurobond issue being five times oversubscribed. By August 2022, however, inflation was at 31% and the Ghanian cedi had lost 36% of its value. The rating agencies Moody’s and Fitch downgraded its Eurobonds to ‘junk’ – effectively shutting Ghana out of international credit markets. The Central Bank responded by hiking interest rates by more than 850 basis points (8.5%) between November 2021 and August 2022, resulting in Ghana requesting an IMF bailout […]. Yields on its Eurobonds are now over 21%.

In short, CRAs aren’t immune to the general tendency towards unreasonable catastrophizing about African economies during cyclical downturns. And because of their signaling power to market players, their predictions are often self-fulfilling.

III: Thinking beyond the CRAs: the domestic origins of African sovereigns’ information problem

An outstanding challenge not addressed by standard critiques of CRAs is the information problem posed by the structure of African economies (largely informal and commodity dependent) and regional government’s lack of policy autonomy (too many governments in the region act merely as implementation arms of foreign policy entrepreneurs and lack their own coherent and productive developmentalist policies).

The cascade of pro-cyclical downgrades witnessed in the region in the last four years are understandable in a world where African states’ economic performance closely tracks global commodity prices — which is true for the region. At the same time, high levels of informality increase the cost of revenue mobilization and limit governments’ capacity to service their debts. Therefore, it shouldn’t come as a surprise that commodity dependence and the illegibility of informal sectors perhaps lead CRAs to attach more weight to global economic cycles (especially when commodity exports are the biggest sources of government revenues).

In the same vein, the outsize influence of the World Bank, the International Monetary Fund (IMF) and other foreign actors in shaping policy in African capitals muddles informational signals regarding fiscal stability in individual countries. For all the talk about “strong democratic institutions,” foreign engagements with African states still rely a great deal on personal relationships with individual presidents and ministers. Under the circumstances, only a foolish rating analyst would base their decisions on the formal rules of public finance management or the roles of institutions like legislatures and courts in protecting creditors’ property rights.

The Bank and the IMF, in particular, often serve an important “certification” function with regard to African credit worthiness. Yet they, too, have the same informational limitations as CRAs, not to mention their lack of capacity to influence political decision-making in African capitals. This reality enables politicians to arbitrage positive signals from the Bank and the IMF, while avoiding the disciplining forces of markets — including regarding budget transparency, investments in fiscal capacity, the prudent utilization of borrowed money, and robust public finance management mechanisms. Furthermore, the lack of depth in regional financial markets limits potential informational and political influences on African policymakers by politically-connected domestic creditors.

Simply stated, there’s a supply side component to the information problems occasioned by CRAs’ lack of investment in knowing individual African economies.

In order to improve on the current situation, African governments and CRAs need to sharpen the signals sent to credit markets. On their part, African governments should invest in transparent budgeting processes (from conception through implementation), predictable policymaking, and macroeconomic conditions that are conducive to mass creation in the “formal” sector. It is also on them to get intimately acquainted with and “game” the CRAs’ input variables, especially in anticipation of global cyclical downturns. And above all, credible domestic policy autonomy would reduce the noise introduced by foreign policy hawkers in their creditworthiness signals.

Clarity of information signals coming from African governments would make it easier for CRAs to develop country-specific knowledge and calibrate both their quantitative and qualitative models appropriately. Yet even before that, CRAs ought to heed their critics’ call to improve their regional presence and country-specific knowledge, if only to make their models and processes more legible (and therefore influential) among African policymakers and private sector actors.

When all is said and done, building credibility will take time and experience; and largely depend on the track records of African sovereigns. There is no short cut around sound PFM practices and consistent commitment to policies that promote the growth of export-oriented firms and mass job creation. Pressuring CRAs to reform or relying on the Bank and IMF for certification may help, but only experience and well-documented success stories will convince investors to bet on African economies.