Optasia, an artificial-intelligence-based lending platform backed by major South African banks, is expanding into Ethiopia and Egypt, CEO Salvador Anglada told Semafor. The company is betting that its machine-learning credit models can keep default rates to a fraction of those seen at traditional commercial banks.
Putting a lens on Africa’s informal credit flows
Optasia, which went public last year, uses mobile data and AI to deliver micro-loans, working capital and airtime advances to unbanked consumers and small businesses. Its operations provide insight into how capital circulates in Africa’s large informal economy — an area often missed by standard economic reporting on poverty, employment and GDP.
Key figures
- Optasia facilitated about $6 billion in credit across 38 developing-market countries last year.
- The group-wide default rate stands at 1.2%, which corresponds to roughly $60–70 million in losses.
- The current expansion is in early stages; the target markets have a combined population exceeding 200 million.
Why Ethiopia and Egypt?
Bank credit to the private sector is among the lowest globally in Ethiopia — under 10% of GDP — after decades in which state-led financing focused heavily on public infrastructure. In 2024 Addis Ababa opened its banking sector to foreign investors for the first time in 50 years to attract international capital to a historically closed, credit-starved market of about 100 million people.
In Egypt, bank credit to the private sector is around 30% of GDP, with lenders directing capital to government debt and blue-chip corporates. As a result, the loan-to-deposit ratio remains unusually low, at just above 50%, according to World Bank data.
AI, scale and risk management
Optasia began 14 years ago as a tool to sell on credit; today direct cash and digital loans represent about 75% of its transaction volumes. The company runs tens of millions of daily credit decisions through a technology stack of more than 200 machine-learning models, which draw on thousands of alternative signals per user.
Anglada says traditional banks cannot profitably service $20 or $50 micro advances because they are not built on digital-native systems and lack the complex algorithms required. Optasia’s AI scoring enables lending to customers with no formal credit history at risk levels lower than many retail banks. Informal enterprises, such as street vendors and local shopkeepers, account for roughly one-third of total credit volumes.
Ownership and partnerships
South Africa’s FirstRand holds a significant equity stake in Optasia, giving the bank exposure to the platform’s 400 million user base without expanding branch networks or incurring manual know-your-customer overheads. According to Anglada, FirstRand views Optasia as a specialized technology and platform provider that reaches the base of the pyramid in ways a traditional bank cannot.
Chronos Capital is the largest shareholder with a 30% stake, while FirstRand holds 26%.
The stakes
Anglada expects the roughly 1.2% loss rate to hold even as the company scales into densely populated markets with a combined population above 200 million. If that holds true, the low default ratio would make Optasia’s business model viable in environments where traditional banks often fail: unsecured lending portfolios at sub-Saharan African banks typically write off 10–15% of loans as bad debt.
Related observation
Former Kenyan ICT Minister Joe Mucheru has noted that machine-learning credit scoring is accelerating micro-loan growth in sub-Saharan Africa. Data from fintech platform JUMO indicate that more than 50% of informal micro-enterprises receiving algorithmic loans hire at least one additional employee.
