2 Min Read

At some point in your accounting studies, you learned how credit assessment works. Bureau report. Payslip. Employment verification. Ratio analysis. Tick the box.

Optasia's first JSE results, published this week, show what happens when none of that applies.

The AI-powered fintech grew revenue 75% to R4.47 billion. Its micro-lending business was up 149% year-on-year and now accounts for 63% of group income. It operates across 38 countries, distributes through MTN and Vodacom locally, and has extended more than $20 billion in small loans to date. FirstRand holds a 20.1% stake.

If you were reading us back in October, this one should look familiar. We covered FirstRand's decision to acquire a 20.1% stake in Optasia ahead of its JSE listing

South Africa has roughly 10 million unbanked adults. Unemployment sits at 32%. Close to two-thirds of the population lives below the poverty line, and a big portion of borrowing here does not cover luxuries. It covers groceries. Electricity. School fees. Transport to work.

Credit demand in this country is not optional for a lot of people. It is baked into how households manage cash flow between income cycles. But traditional banks have mostly failed this segment because their risk models need things that simply do not exist for people outside the formal economy: payslips, credit histories, consistent employment records.

The South African alternative lending market is projected to grow from around $1.6 billion to close to $10 billion by 2033. That growth is being driven by AI-powered platforms that do not need a payslip. They need data. And mobile-connected South Africans generate a lot of it.

Optasia's model uses 5,000+ data points to make a credit decision in real time: mobile usage, airtime purchase patterns, transaction behaviour, top-up frequency. The company runs this through MTN and Vodacom, two networks that between them cover most of the country. If you have a SIM card and a phone, you have a credit profile in Optasia's model.

The company started in airtime credit, lending people small amounts to cover a call or a data bundle when funds ran low. This year, its micro-financing solutions business grew 149% and now accounts for 63% of group revenue, overtaking airtime credit for the first time. Revenue came in at R4.47 billion, up 75%. Operating profit hit R1.2 billion. Net profit was up 19% to R725 million, though once-off IPO costs took some of the shine off the earnings line.

The model is working because the market need is real and the distribution, through mobile networks, already exists at scale.

Under IFRS 9, this creates a challenge for future auditors. How do you test an expected credit loss provision when the model behind it is a machine learning algorithm trained on mobile behavioural data? What does sufficient audit evidence look like when you cannot manually trace the inputs?

South Africa's credit market is large, underserved, and being reshaped right now. You will audit, advise, or work alongside businesses doing exactly this. Knowing how AI credit models work, and how to interrogate them, moves from theory to practice faster than most students expect.

Until next week,
The Journal Entry Team

The journey to those four letters CA(SA), is a long one, let’s not kid ourselves. We all know the high volume of coffee and library hours required to get there.

The last thing you want, is to spend your first working years at a firm that you don’t enjoy. You want to work somewhere that aligns with your specific skillset, because working in an environment you actually enjoy is important.

Aptihire helps students discover, apply for, and land their dream SAICA, SAIT & CIMA articles.

It is a free tool for students. It considers your unique personality and skills to provide tailored recommendations. This platform ensures you find a workplace where you can successfully turn your academic theory into practice.

Explore Aptihire here

If you enjoy our content we’d love it if you showed your friends and peers!

Please copy and paste this link to others:

Work With Us