South Africa's Poorest Are Borrowing to Eat. The Iran War Just Made It Worse.
What the banks' best quarter in years is hiding — and why April 2026 changes everything
Some background first
I was appointed as an economic reporter for Finweek a few weeks before the rand manipulation by a bunch of traders exploiting thin trading volumes in December 2001. This led to an a formal inquiry into the rand and a surge in inflation over 2002. I recognised back then - armed with strong theoretical knowledge at the time as a new Econometrics / Economics graduate - that our monetary policy and our policy tools were nonsensical. I lament this often. I understand the driving pressures dictating the SARB’s mandate targets inflation, but its main tool interest rates, is a blunt tool that responds to inflation drivers that often has nothing to do with internal factors. Now inflation will surge, the SARB will respond in kind and who knows how long the war illegal war with Iran will carry on for.
The TransUnion Q4 2025 Industry Insights Report arrived at a curious moment in history. Its data captures the last clean quarter before the world changed. By the time it was compiled, the Strait of Hormuz had been effectively closed, Brent crude had surged past $100 a barrel, the rand had shed more than 6% of its value, and the SARB’s carefully calibrated rate-cutting cycle had been placed on indefinite hold. The Q4 data is a photograph of a hospital patient in recovery - just as someone handed them a new diagnosis.
Read with that in mind, the report is not just a quarterly scorecard. TransUnion’s report is a baseline against what comes next.
Ubiquity AI deals with consumer debt. The global economy has never seen a confluence of such unpredictable forces. I have never witnessed such complexity in the global economy. I’ll write about that on another day.
More urgently, I’ve pondering the consequences of the war on South Africa. It took some time to work through what the data actually shows - product by product, risk tier by risk tier - and then put it in the context of where we are today. And I’d strongly recommend reading the TansUnion report as a fantastic starting point.
The Scale of South African Consumer Credit
First, the landscape. South African consumer credit is dominated by a handful of large products. Home loans alone represent over R1.27 trillion in outstanding balances. Vehicle finance - over R584 billion. Bank personal loans - nearly R304 billion. Together, these three products constitute approximately 87% of all formal consumer credit outstanding. Credit cards, at R195 billion, and the sprawling non-bank personal loan market, at R93.8 billion, complete the picture for secured and semi-secured lending. Retail credit - clothing accounts, revolving, and instalments - is large in account numbers but modest in rand terms, its importance lying not in balance size but in its role as the credit safety valve for lower-income South Africans.
Who Currently Gets Credit, and Who Doesn’t
Here is where the report’s most important structural signal lives - one that requires reading between the lines.
The headline numbers for originations look broadly healthy. Bank personal loan originations grew 10.2% year on year. Vehicle finance originations rose 9.9%. Home loan originations jumped 8.5%. Credit cards grew 8.0%. At first glance, this looks like a recovering credit market distributing opportunity more broadly.
Look closer.
Average new credit card limits fell 10.0% year on year - with the sharpest cuts at the subprime and near-prime end. That same market segment now accounts for 58% of all open credit card accounts. Lenders are growing the account base but deliberately shrinking per-account exposure for higher-risk borrowers.
In vehicle finance, the entire growth in loan values is driven by a shift toward new vehicles (partly incentive-driven), while the average new loan amount grew only modestly.
In home loans, the average new credit line rose 4.8% to R964,700 - growth concentrated at the top of the affordability spectrum. Under IFRS 9, high-risk borrower segments could find credit more expensive or harder to access, as the upfront recognition of expected losses makes such lending less attractive from a profitability and capital perspective.
This is IFRS 9’s fingerprint: Since 2018, South African banks have been required to provision for expected credit losses (ECLs) rather than incurred losses. Under the old IAS 39 standard, banks could extend credit optimistically and only recognise impairment when default actually arrived. Under IFRS 9, banks are required to assume and account for a degree of credit loss for all their assets, estimating losses on each asset for both the coming year and the lifetime of the asset. CFO South Africa The practical consequence: lenders can now better differentiate between borrowers, potentially offering more competitive rates to lower-risk clients while appropriately pricing for higher-risk profiles.
In practice, this means the credit expansion we see in Q4 2025 is quality-tiered. Banks are growing their books - but their growth is based on leaning heal on wealthier people, while excluding middle and lower income.
Lenders have invested heavily in sophisticated credit risk models that incorporate sector-specific risks, individual client payment histories, and various macroeconomic scenarios pertinent to South Africa, according to FineIT Blog. When those models are loaded with an environment of sticky structural unemployment, low growth, and now a geopolitical shock, their output systematically disadvantages lower-income borrowers - not through explicit discrimination, but through the cold arithmetic of expected loss.
The non-bank personal loan market tells the counterpoint to this story. Account volumes surged 31.2% year on year. Originations grew 14.7%. But average account balances fell 20.2% to just R10,000 — small-ticket, short-cycle lending that banks won’t touch. This is not financial inclusion in a celebratory sense. It is the market routing around bank conservatism to serve borrowers who have nowhere else to go. And as the report notes, balance-level delinquency in non-bank personal loans rose a striking 560 basis points year on year. The stress isn’t spreading to more borrowers - it’s concentrating among those already struggling.
The Delinquency Picture: Recovery Is Real, But Uneven
The most genuinely encouraging signal in the report is the improvement in delinquency across most secured products. Vehicle finance delinquency fell 59 basis points year on year to 6.8% at the account level.
Bank personal loan consumer-level delinquency dropped a remarkable 366 basis points. These are not trivial moves - they reflect a real improvement in borrower quality and the SARB’s easing cycle beginning to reduce household debt service burdens.
But zoom out and the full picture is more complex. Non-bank personal loans carry a 52.9% consumer-level delinquency rate. Clothing accounts sit at 36.7% consumer-level delinquency.
Retail instalments: 27.2%.
The credit market has a dual structure, and it has had one for some time - a formal, bank-led segment with manageable delinquency, and an informal, high-volume segment where more than half of all consumers are behind on payments at any given time.
This structural divide raises a measurement question that regulators should be asking more loudly: when we say “financial inclusion,” which segment are we describing? Including more consumers in the non-bank loan market - at R10,000 per ticket with a 53% balance delinquency rate — is not the same kind of inclusion as extending a home loan or a consolidation loan that improves balance sheet stability.
Non-Discretionary Credit: When Basics Become Borrowed
One of the most important - and least remarked upon - dynamics in this report is the character of what South Africans are actually using credit for. Look past the label “clothing accounts” and consider what the data actually implies. These are 18.5 million accounts, carrying R43.6 billion in outstanding balances, used primarily for everyday household expenditure: school uniforms, winter clothing, basic household goods. The account base grew 5.8% year on year. More than 10 million consumers carry a balance.
This is non-discretionary credit. Not aspirational purchasing, not investment. Basic consumption, financed on revolving terms, with a 36.7% consumer-level delinquency rate.
The BNPL data compounds this picture. The report notes that 57% of South Africans now hold a buy-now-pay-later product, and 36% have used one multiple times in the past year.
This is displacing formal retail instalment credit (down 19.4% in originations), but is it actually improving consumers’ financial position? BNPL typically charges zero interest to the consumer but accrues risk to the provider - and more importantly, it enables a pattern of consumption that may be ill-suited to income volatility.
The composite picture: South Africans at the lower end of the income distribution are increasingly using short-cycle, high-frequency debt instruments - non-bank personal loans at R10K average, BNPL, clothing accounts - to finance consumption that was previously funded from wages. This is not irrational behaviour in a stagnant real income environment. It is, however, structurally fragile in ways that are about to be stress-tested.
The Iran Shock: From the Strait of Hormuz to Your Grocery Bill
Everything above describes a Q4 2025 that was cautiously improving. Now we need to describe where that market is today, in late March 2026.
Dr. Fatih Birol, the Executive Director of the International Energy Agency (IEA) described the situation caused by the war as the “greatest global energy security challenge in history.” Brent crude oil prices jumped about 15% in the opening days of the conflict, then surged to $120 a barrel as it deepened and the market began pricing in the risk of sustained disruption.
South Africa’s central bank maintained its main lending rate at 6.75% last Thursday, saying caution was needed as higher energy prices triggered by the U.S.-Israel war against Iran would push up inflation. The SARB, which had been widely expected to cut rates again in March, held unanimously.
Governor of the SARB, Lesetja Kganyago, said the central bank’s projection model showed rates unchanged for a longer period, postponing cuts seen in January.
The inflation numbers tell the story. Headline inflation is projected to rise to around 4% in the second quarter, led by fuel inflation exceeding 18%, before gradually easing back to 3% by late next year under the baseline forecast. According to Trading Economics that is the baseline scenario.
Independent economist Elise Kruger said the fuel price increases on April 1 “will be the highest ever to be implemented in a single month in South Africa and will likely derail the fragile economic recovery envisaged for the country in 2026.”
According to Business Day, South Africa imports about 80% of its fertiliser, which accounts for roughly 30% of agricultural input costs. Daily Maverick Fertiliser prices have risen roughly 30% amid supply disruptions linked to the closure of the Strait of Hormuz. This is a primary agricultural input cost shock with a 6–12 month lag to food price inflation, which means that even if the conflict were to end tomorrow, food price pressure would persist well into H2 2026.
The price of diesel is the lifeblood of South Africa’s distribution economy gives me heartburn. Diesel price increases feed directly into the cost of moving every item from manufacturer to shelf. The gains which were achieved through the gradual reduction of the basic fuel price during 2025 will be erased and the consumer will, inevitably, begin to feel this change in increasing prices at the till.
According to StatsSA: Fuel, which includes petrol and diesel (crude oil derivatives), is categorised under Transport.
Group Weight: The weighting for fuel has decreased to 3.89% in the updated basket, down from 4.82%.
Secondary Weighting (Within Transport): Fuel remains the largest single component of the Transport category, accounting for roughly 28% of the total Transport weight (which is 13.9%).
What the Iran Shock Does to the Credit Portfolio We Just Described
Let me be specific about the transmission mechanism into the credit data we have just analysed.
Clothing accounts and retail credit will feel the impact first and hardest. These consumers - already showing 36.7% consumer-level delinquency on a baseline - will see their non-discretionary expenditure costs rise significantly: food, transport, utilities.
The April 1 Eskom tariff hike of 8.76% compounds the fuel shock. Real disposable income for this segment will contract.
The most likely outcome: higher utilisation rates, faster roll-through to delinquency, and growing reliance on BNPL and non-bank personal loans to cover the gap. The 560 basis point rise in non-bank balance delinquency that was already appearing in Q4 - before the shock - is almost certain to accelerate.
Non-bank personal loans are the most acutely exposed segment. At R10,000 average balances with 52.9% consumer-level delinquency already, this market was operating at the edge of sustainability in Q4. A 4–5% CPI environment with fuel inflation over 18% will reduce borrower repayment capacity for millions of people who are already missing payments on more than half of all balances in this market.
If lenders respond by tightening origination standards - which IFRS 9 forward-looking models will essentially force them to do, as macroeconomic scenarios are updated - access for the most vulnerable borrowers contracts precisely when need peaks.
Vehicle finance carries the hidden time bomb of 56.4% of loans at 72+ months. At a prime rate of 10.25%, a 72-month vehicle loan is already carrying a meaningful real interest burden. If the SARB is ultimately forced to hike - which its own adverse scenario modelling now contemplates - the monthly servicing cost on these long-dated loans increases, and negative equity windows extend. Lenders with strong residual value forecasting are better placed; those who approved long-dated loans at incentive-boosted transaction prices may face portfolio stress.
Home loans are the most insulated segment in the short term. The prime-linked mortgage market benefits from consumers who, almost by definition, had to pass a rigorous affordability test to qualify. First-time buyers at 51% of new originations are more exposed - a rate reversal mid-mortgage is highly stressful for this cohort - but home loan delinquency entering this shock at 7.5% is structurally sound.
Credit cards carry the worry of already-rising delinquency (balance-level up 71 basis points year on year) in a market where 58% of accounts are subprime. The revolve-and-carry dynamic means that for the consumers already stressed, cost-of-living increases will deepen utilisation and pull more balances into the 3+ MIA bucket.
The Rate Environment Is Broken (For Now)
Inflation forecasts were raised to 3.7% for 2026 (from 3.3%) and to 3.3% for 2027. The central bank also revised its policy outlook, now projecting only one rate cut instead of two previously, while assessing two possible Iran conflict scenarios - a short-term two-month scenario and a prolonged one-year scenario, both implying the need for higher interest rates, according to Trading Economics.
This matters enormously for the credit market. The Q4 2025 recovery story was premised on a continued rate-cutting cycle. The SARB had cut 150 basis points from the 2024 peak. Vehicle finance, home loans, personal loans - all benefited from the improved affordability environment this created. That cycle has now been suspended indefinitely.
Kganyago outlined two risk scenarios - both pointing to possible interest rate hikes. A shorter conflict could result in one hike, while a prolonged war with sustained high oil prices and a weaker rand could require multiple increases. In both cases, inflation is expected to rise above target before gradually easing back to 3% over the next two years. The South African
The anticipated increase in inflation, according to Daily Maverick, has likely removed the prospect of more interest rate cuts in 2026, with rising inflation eroding real disposable income, reducing household purchasing power. I agree.
The SARB finds itself in the position described astutely by one commentator as fighting the wrong war with the right weapons. The SARB is fighting a 21st-century mess of geopolitical blockades and failing state infrastructure with nothing but basis points and press statements.
Raising rates to contain fuel-driven inflation doesn’t produce more oil; it simply transfers the pain from the price level to the interest rate channel. Households experience both.
Ubiquity AI’s Forward View: What Prudent Analysis Demands
South Africa enters Q1 2026 with a consumer credit market that showed genuine signs of stabilisation in Q4 2025 - and is now absorbing a shock that the Q4 data could not anticipate.
Several things follow from a rigorous reading of both datasets together.
For banks and regulated lenders, IFRS 9’s ECL models will be updating macroeconomic scenarios in real time as inflation forecasts rise.
This mechanically increases Stage 2 and Stage 3 provisioning requirements across portfolios with meaningful exposure to lower-income and higher-risk borrowers.
Expect reported impairment charges to rise in H1 2026 results, which in turn restricts the capital available for new credit extension to precisely the segments that most need it. This is the procyclicality of accounting-driven provisioning in practice - well-documented in academic literature, and now about to play out again.
For the non-bank sector, the business model faces acute pressure. Origination growth of 14.7% into a market already carrying 52.9% consumer delinquency, with balance-level delinquency up 560 basis points year on year, suggests that returns on the marginal loan are thinning rapidly.
A cost-push inflation shock that reduces borrower repayment capacity will either force higher pricing - which faces National Credit Act constraints - or higher default rates. Many non-bank lenders will be forced to reduce origination volumes, which means the credit safety valve for lower-income South Africans starts to close.
For the SARB, the asymmetric risk is particularly uncomfortable. A supply-side inflation shock is the worst kind to manage with a demand-side tool. Rate hikes cannot produce oil; they can only reduce demand, which at current South African income levels means tipping already-fragile household budgets into default. The SARB’s own adverse scenario analysis - inflation above 5%, rate hikes required, target missed through 2027 - is not a tail risk. It is an entirely plausible central scenario if the conflict endures beyond six months.
For policymakers broadly, the Q4 data on non-discretionary credit spending should be read as a structural alarm. When 18.5 million clothing accounts carrying R43.6 billion in balances at 36.7% consumer delinquency represent the primary credit access mechanism for lower-income South Africans, the social stability implications of further real income compression are not marginal.
Credit is functioning as a social buffer that wages and the fiscal transfer system are no longer adequate to provide.
That buffer has a breaking point.
Bottom Line
The TransUnion Q4 2025 report described a credit market in cautious, quality-tiered recovery. Delinquency was improving where it mattered most. Lenders were deploying capital with discipline. Consumers were prioritising repayment where they could.
What it did not describe - because it could not - was a market about to absorb April fuel increases described as the highest ever implemented in a single month, a rand more than 6% weaker, fertiliser costs up 30%, food price inflation building a 6–12 month lag, and an interest rate environment that has been locked in place indefinitely.
The Q4 2025 data is the last photograph of the patient before the diagnosis. The next quarterly report - Q1 2026, due sometime mid-year - will be the first picture of how the patient is responding to treatment. The prognosis depends almost entirely on something South African monetary policy has no influence over whatsoever: how long the United States and Israel continue their campaign against Iran, and what happens to the 20 million barrels of daily oil flow that once passed through the Strait of Hormuz.
South Africa’s credit market is, in the most literal sense, hostage to a geopolitical variable twelve thousand kilometres away. The Q4 data told us we had built some resilience. We are about to find out how much.
Data sourced from TransUnion South Africa Q4 2025 Industry Insights Report. Macroeconomic context sourced from SARB MPC statements (March 2026), Reuters, Business Day, Daily Maverick, IOL and Engineering News.
This analysis represents the independent views of the Kaveer Beharee, Founder of Ubiquity AI https://ubiquity-ai.com and does not constitute financial, credit, or investment advice.
#ECL #IFRS9 #SARB #Iran #war #inflation #transunion #credit #lending #lending #poverty #financialinclusion