South Africa’s political risk is, by most measures of domestic governance, considerably greater than the bond market currently admits. The central accountability question raised in recent analysis published by The Common Sense, at https://www.thecommonsense.co.za/Editorials/south-africa-s-political-risk-greater-than-bond-market, is whether the institutions responsible for the country’s fiscal and political management have genuinely earned the confidence that market pricing now reflects, or whether investors are extrapolating from a governance record that no longer exists.
The trigger for that question is a market signal. The yield differential between South Africa’s 10-year government bond and the equivalent United States Treasury bond has narrowed to levels last seen between 2004 and 2007, a marked improvement in the market’s assessment of South African assets since the formation of the Government of National Unity (GNU) after the May 2024 election. The differential is a broad indicator of the additional yield investors demand for holding South African rather than American government debt, and because the two bonds are denominated in different currencies, it also reflects relative inflation, interest rate and exchange rate expectations. Its compression therefore matters: it is a judgment, however implicit, that the country’s governing arrangements have become more durable and its public finances more secure.
The historical record explains why that judgment deserves scrutiny. Between 1994 and roughly 2007 the differential fell steadily, and between 2004 and 2007 the conditions underpinning the pricing were demonstrably present. The economy grew at approximately 5% annually. Government debt had been halved from the levels inherited in the 1990s. The government moved into a budget surplus, and the African National Congress (ANC) exercised, at least superficially, almost complete political control under a predictable fiscal and economic policy direction. After 2008 the differential rose and held a few points above that earlier range, spiking on successive risk events and culminating in the Covid-19 shock. When the GNU was formed, the spread broke well below the post-2008 average to levels not seen since the mid-2000s.
The problem, as the analysis sets out, is that the country’s underlying circumstances bear little resemblance to 2007. Economic growth is barely above 1%. Government debt stands near 80% of GDP. The budget remains in deficit. And the political environment is deeply fragmented: the ANC is losing electoral support, its principal coalition partner is increasingly conscious of the political costs of remaining in government, and two radical opposition parties have secured sufficient support to threaten the existing national settlement.
The market’s optimistic pricing, the analysis argues, rests on three assumptions: that further ANC electoral weakness will be positive for South Africa, that the GNU is durable enough to withstand the pressure following the November local government elections, and that the National Treasury has restored a sustainable fiscal path. Each, it contends, deserves considerably more scrutiny than it has received.
The first and most immediate test is the local government election scheduled for 4 November. Research published by The Common Sense suggests that between 100 and 130 municipalities could emerge without a governing majority, compared with roughly 70 after the previous election, an unprecedented degree of fragmentation at municipal level requiring an equally unprecedented series of coalition negotiations. In scores of councils the ANC and the Democratic Alliance (DA) may collectively command a majority, but no comprehensive agreement exists between the parties on how such governments should be formed. Without one, the election could produce chaotic bargaining, failed coalition negotiations and unstable municipal administrations. The precedents set will reach national politics: where the ANC and DA cooperate, the logic of the GNU is reinforced; where either governs with other partners, the GNU is exposed as merely one of several possible arrangements.
Neither governing party, moreover, is unequivocally committed to the arrangement. The DA must demonstrate to voters that participation in national government produces meaningful policy change while avoiding responsibility for the ANC’s continuing failures, and The Common Sense judges its current leadership at best ambivalent about the arrangement, quite unlike the previous leadership. The ANC, for its part, is becoming progressively more dependent on the DA even as powerful constituencies within the party remain hostile to the required policy concessions, and its electoral strategy depends on persuading traditional supporters that it has not surrendered control to its principal historical opponent. The result is an uneasy arrangement in which both parties are finding increasingly compelling reasons to distance themselves from one another, a tension the November elections will make impossible to defer.
Compounding this is the weakening political position of President Cyril Ramaphosa, whose authority has been gravely damaged by the continuing Phala Phala controversy, the ANC’s declining electoral support, and the failure of the economy to deliver the expected post-GNU recovery. Polling data cited by the publication indicate that the greatest share of voters believe his removal from the presidency and the ANC leadership would make no difference to their lives or would improve them. The publication suspects removal schemes are under consideration, with no firm successor apparent. Patrice Motsepe is the closest thing to a positive answer in its polling data, though no decision has been taken on whether he would run.
Meanwhile the radical parties have acquired substantial leverage. The uMkhonto weSizwe Party has displaced the ANC across significant parts of the electorate in KwaZulu-Natal and parts of Gauteng and southern Mpumalanga, while the Economic Freedom Fighters, through cooperation with the ANC and the MKP in municipal governments, has helped establish an alternative governing model that could appeal to ANC factions hostile to the national GNU.
The fiscal record raises parallel doubts. Last week’s Medium-Term Budget Policy Statement, delivered by Finance Minister Enoch Godongwana, announced that government debt had peaked at 78.4% of GDP, below the 78.9% previously forecast, and that the current-year deficit would narrow to 3.9% of GDP against the 4.0% projected in February. But a deeper analysis by The Common Sense found the headline numbers had been “engineered”: government expenditure was revised upwards from 32.6% to 33.4% of GDP while revenue was raised from 28.6% to 29.5%, the larger revenue increase being what allows the Treasury to present a marginally smaller deficit despite planning for higher spending.
None of this guarantees an immediate repricing; global capital flows, inflation expectations and movements in US Treasury yields can move the spread independently. But the domestic case for mid-2000s-calibre risk assessment is difficult to sustain. The publication’s scenario work anticipates a moderate repricing over the next 12 to 18 months on its base case, with a sharp, several-point widening under the downside scenario. Whether the institutions now holding power can withstand that test, or whether the November elections force the question sooner, remains the open question hanging over the GNU.