Funding for microlending businesses in South Africa

The two capital questions every microlender faces

Book funding and business funding are different instruments, priced by different funders, secured differently — and conflating them is the most common structuring error in the sector. Book funding is debt: facilities secured against the receivables book, advanced at a rate to the book’s collectable value, priced off demonstrated collections performance. Business funding is equity or mezzanine into the operating company: origination technology, branch or agent networks, compliance capacity, and the working buffer regulation requires. A lender that raises equity to fund book growth dilutes itself unnecessarily; one that stretches a book facility to fund operations breaches covenants at the first collections wobble. The split — decided before any funder conversation — is the raise.

The layered book-funding structure that actually works

Institutional book funding in South Africa is built in layers. Senior facilities — from banks’ wholesale desks and larger private credit funds — advance at conservative rates against the book (advance rates set off vintage performance, typically with portfolio-at-risk covenants). Mezzanine or first-loss layers — from specialist credit funds, impact debt investors and DFIs — sit beneath the senior and are what unlock senior leverage at all for younger lenders; without a subordinated layer, most senior lenders will not engage below significant scale. Equity tops the stack for the operating business. The layering is not optional sophistication: it is how every institutionally funded lender in the market is built, and arriving with the structure already designed signals competence before the data pack is opened. Caban structures the mezzanine layer as a core practice.

The data pack funders actually interrogate

Five exhibits decide microlending diligence. Vintage curves by cohort — how each month’s originations performed over time; blended averages that smooth deteriorating cohorts are the classic concealment and the classic discovery. Roll rates — the movement of accounts through arrears buckets, which is where collections capability shows or doesn’t. Portfolio-at-risk (PAR30/PAR90) against provisioning policy — optimistic provisioning versus actual write-offs is found in every diligence that fails. NCR standing — registration class, compliance history, and affordability-assessment practice under the National Credit Act; a compliance finding mid-raise is fatal. Collections infrastructure — payroll deduction, debit-order management, field capability — because funders price the mechanism, not the intention. A lender with honest cohort data and a defensible niche (payroll-deducted lending, merchant cash advance against card turnover, secured micro-asset finance) can raise in any market; a lender with blended numbers cannot raise in the best one.

Who funds South African lenders

Bank wholesale and structured-finance desks for established books; local and international private credit funds across senior and mezzanine; impact debt investors and DFIs — for whom SME and inclusive credit is an explicit mandate, particularly where the book serves under-banked segments the IFC-estimated credit gap describes; and equity from specialist financial-services investors for the platform itself. Each reads the same data pack differently: DFIs weight development outcomes and compliance, private credit weights structure and covenants, banks weight scale and track record. A process run across all of them in parallel — with the structure pre-designed — is how competitive terms emerge.

Caban’s track record and how to start

Microlending is one of the sectors in Caban’s own transaction track record. The work spans designing the book/business split, building the cohort-level data pack to institutional standard, structuring the senior/mezzanine/equity stack, and running the raise across banks, private credit, impact and DFI capital in parallel. If the book is performing and the data is honest, the sector is more fundable than most founders believe — the constraint is almost always preparation, not appetite. Start with the readiness check, or the broader lending-business funding guide if the model extends beyond micro-credit.

The questions microlending funders will ask

• Vintage curves by monthly cohort — raw, not smoothed.
• Roll rates by bucket, with collections actions mapped to each.
• PAR30 and PAR90 against provisioning policy — and against actual write-offs.
• NCR registration class, compliance history, and the affordability-assessment file.
• Funding cost today, layer by layer — and margin at the NCA rate caps.
• Collections mechanism: payroll deduction, debit order success rates, field capability.
• Concentration: employer, region, channel — and what breaks first in a downturn?

A lender with these exhibits indexed raises in weeks; without them, not at all.

Finally, model the margin at the caps: NCA rate ceilings mean funding cost passes through to viability directly, and every funder will test whether the book’s economics survive senior pricing plus a stressed cost of collections inside the regulated maximums. Arriving with that arithmetic already run — honestly, at today’s rates — is the difference between negotiating structure and defending viability.

And retest the arithmetic quarterly as funding costs move — the margin that worked at last year’s rates may not survive this year’s, and discovering that in a covenant conversation is the expensive way to learn it.