🏦Optasia Eyes Banks Next. Investec Restructures.

Afternoon, Elly here. Elon Musk’s Starlink tried to enter Namibia but they said no, no, no. This week: a Reserve Bank study revealed that open banking - the initiative that’s supposed to help the financially excluded - is actually excluding them. Investec just reorganised to chase a R600 billion opportunity where competitors are making 50% returns. And Optasia, the microlending fintech that taught telcos how to lend now wants to sell the same playbook to banks. Let's get into it.

In other news, if you enjoy The Banking Brief, please whitelist this email and move it to your primary inbox. And if you were forwarded this email: sign up here.

Inside:

🔓 Reserve Bank research finds South Africa's market-driven open banking framework excludes rural poor, Eastern Cape residents, and credit-thin users - the exact people it was meant to serve.

💰 Investec restructured to chase R600bn mid-market lending, targeting 10,000 clients by 2030 in a segment where Standard Bank extracts 58% return on equity.

📱 Telco-to-bank infrastructure play: Optasia, which taught MTN and Vodacom to lend using airtime patterns, now pitches the same model to banks for 15-20 million credit-invisible South Africans.

BANKING

Open banking promised inclusion

What’s going on? A Reserve Bank working paper published this week revealed South Africa's open banking system has created an ecosystem that bypasses the people it was designed to help. Open banking works through secure APIs - application programming interfaces, basically digital pipes that let different software systems talk to each other - allowing your favourite startups like Yoco and Lulalend to offer services without holding expensive banking licences. The research team analysed who's actually using these services and found women make up 59% of users, which sounds good until you dig deeper. Gauteng and Western Cape dominate adoption. Meanwhile, Eastern Cape, KwaZulu-Natal, and Limpopo are getting left behind.

So what? The researchers recommend South Africa shift to a Reserve Bank-led or hybrid model with mandatory data-sharing standards, consumer consent protections, and parallel investment in digital literacy and connectivity. Moving to mandatory participation frameworks would force banks to open their APIs to competitors, removing the lever banks currently use to control which fintechs access their infrastructure and at what price. For fintechs building on voluntary bank partnerships, this shift could make or break business models. If regulators mandate access with price controls, platforms charging merchants for conversions might see their margins compress. Meanwhile, the Kenya-Rwanda licence passport signed last week creates network effects that reward scale, favouring well-capitalised platforms that can absorb multi-jurisdictional compliance costs.

BANKING

Investec goes after mid-market banking

Investec restructured its South African operations into three divisions, creating a business and commercial banking unit under Nick Riley to target companies earning R30 million to R1.5 billion annually. The segment's a goldmine. Investec's analysis shows incumbents are extracting up to 50% return on equity - ROE, which measures how much profit a bank generates for every rand of shareholder capital invested. Standard Bank reported 58% ROE from this division in 2025. That kind of profitability usually signals underserved customers paying premium prices. The plan: grow from 3,000 clients to 10,000 by 2030, lifting revenue from R1.7 billion to R3.8 billion.

How will the new unit operate? Riley's unit will bundle lending, private banking, and advisory under one roof - a one-stop shop designed to pry businesses away from FNB (market leader), Nedbank (which hired FNB's Business Banking CEO Andiswa Bata earlier this year targeting 25% market share), and Absa. To hit R1.5 billion revenue by 2030, Investec needs about R180 billion in loans outstanding. Acquiring 7,000 new clients means each relationship averages R26 million in facilities. That concentration creates credit risk if the portfolio tilts toward cyclical sectors. The real battle will be fought on technology - mid-market clients need customised credit assessments, treasury services, and cash management. Automating these while maintaining the high-touch service that justifies premium pricing is the operational challenge.

Why now? : incumbents are making so much money they've stopped trying. When Investec enters offering comparable products with better service, FNB and Standard Bank face a choice: maintain pricing discipline or defend market share through relationship investment. Nedbank's move earlier this year suggests Investec isn’t launching a unique strategy - it's synchronised competitive intensity across the segment. But there's a problem with three banks pitching the same prospects with similar promises: differentiation becomes harder and switching rates may disappoint. The 50% returns incumbents have been harvesting suggest they may have been competing on inertia rather than pure service. If you're a mid-market CFO, you're about to get courted hard. The smart play would be to use the competition to negotiate better terms while watching which bank actually delivers on the service promises after you sign.

Optasia: Telcos cracked lending without credit scores. Now they want banks to notice.

What’s happening? Optasia loves been front of mind in the news. Can’t fault them. Optasia spent years teaching MTN and Vodacom how to lend to their subscribers using a pretty simple insight: someone who tops up R50 airtime every Monday for two years without missing a week is probably good for a loan. Traditional credit bureaus miss this because prepaid airtime purchases don't get reported. The telco programmes worked, the loans performed, and Optasia processed billions. Now they're pitching the same playbook to banks. Here's the gap: South Africa has about 24 million credit-active consumers. Another 15-20 million adults are financially active but credit-invisible - they earn income, pay rent, buy electricity, and top up airtime, but have thin or non-existent credit bureau files.

What’s the context? Banks reject them automatically. Not because they're risky, but because traditional scoring models simply can't assess them. Optasia's pitch is that banks already have the transaction data (salary deposits, utility payments, retail purchases) to build alternative scorecards. Telcos proved it works. The challenge here is that banks have built regulatory infrastructure around credit bureau data for decades. Lending outside that framework means explaining to regulators why a loan that scores as subprime on traditional models gets approved. Credit bureaus represent socialised risk across the industry - when a loan defaults after getting a positive bureau score, nobody questions the decision. When a loan defaults based on airtime payment patterns, the bank has to defend that methodology.

What’re the implications? Three scenarios play out. One: A bank or two pilots alternative data programmes targeting gig workers, recent graduates, informal traders. FNB which recently upped its shareholding from 20.1% last November to 26.1% in March will be first in line. Optasia gets limited contracts. Thin-file customers gain access at initially premium pricing. Two: Fintech lenders using alternative data build portfolios from bank-rejected customers. After proving the model, they scale or sell to banks who pay acquisition premiums for capabilities they could have developed internally. Three: The Reserve Bank mandates use of all available creditworthiness data, positioning alternative data as compliance rather than innovation. The telco programmes establish proof of concept. Bank adoption depends on whether competitive pressure, regulatory requirements, or margin opportunities in unbanked markets outweigh the business model tension of expanding access while protecting margins.

Rapid-Fire

🌍 Africa

Rand takes a beating. South Africa's currency hit a four-month low at 17.2275 against the dollar on Monday as Middle East tensions (Trump threatened to 'obliterate' Iran's power plants unless it reopens the Strait of Hormuz) kept oil prices elevated. Analysts expect continued pressure ahead of the Reserve Bank's interest rate decision on 26 March. The central bank held the repo rate - the interest rate at which commercial banks borrow from the SARB - at 6.75% in January with forecasts now seeing it ending 2026 slightly above 6.25%.

Kenya-Rwanda fintech passport signed. The two East African nations signed a licence passporting framework allowing fintechs licensed in one country to operate in the other without duplicate licensing. Details on implementation still emerging, but the mutual recognition agreement creates network effects favouring platforms with multi-market licences.

Starlink blocked in Namibia. Elon Musk's satellite internet service got rejected by Namibia's Communications Regulatory Authority for both a telecoms licence and radio spectrum access. No reason provided in the 23 March government gazette notice, but the regulator can reconsider within 90 days. Starlink operates in several African countries but faces pushback from state telecoms monopolies.

🇺🇸 US

OpenAI says goodbye to Sora. Turns out AI video isn’t as cheap as Sam Altman thinks. Sora’s packing up its bags and ending its mega-partnership with Disney. This week has featured a lot of recanting from OpenAI - ChatGPT shopping, who seem to be struggling with strategy issues while Anthropic, OpenAI’s biggest competitior gains mindshare with fan-favourite rewleases. The AI lab released Sora, its text-to-video model, to ChatGPT Plus and Pro subscribers. The system generates up to 20-second videos from text prompts at 1080p resolution in widescreen, vertical, or square formats. Free tier users don't have access yet.

Startup spotlight

Happy Pay

Cape Town • Founded 2021 • $5M seed (Partech lead)

Happy Pay closed a $5 million seed round to scale Africa's first advertising-subsidised buy-now-pay-later network. The pitch: eliminate interest and fees for consumers by making merchants pay instead - through transaction fees and targeted advertising. Here's how it works: Happy Pay's AI analyses your transaction history, browsing patterns, and purchase timing to predict which products you might buy. When you're shopping, the system embeds instalment payment options directly into the purchase journey and surfaces relevant product ads. Merchants pay when transactions complete. Advertisers pay when their ads convert to sales. You pay nothing in fees or interest.

The founders - Wesley Billett, Patrick Postrehovsky, and Mark Geary - have grown the platform to 600,000+ users. The challenge is finding the sweet spot where advertising density funds zero-interest installments without degrading user experience. Naturally too few ads means insufficient revenue. Too many ads drives users away. Traditional BNPL platforms like PayJustNow and Payflex charge consumers late fees or interest. Happy Pay's bet is that merchant value extraction through higher transaction fees and ad conversions creates better unit economics while expanding addressable market to anyone willing to tolerate ads.

What to watch

On this day in banking history

March 1957: The Treaty of Rome was signed, establishing the European Economic Community - the precursor to the EU. Six founding nations (Belgium, France, Italy, Luxembourg, Netherlands, West Germany) committed to creating a common market with free movement of goods, services, capital, and people. The treaty's economic integration framework eventually led to the euro's introduction in 1999, creating the world's second-most traded currency.

Got feedback? Hit reply. We read everything.