A neutral multi-bank vehicle finance pre-approval layer for South Africa.

1. The problem

South African car buyers approach the dealer F&I desk almost completely blind. The four major lenders publish calculators on their own websites, but each is bank-branded and hides the actual approval likelihood behind opaque “subject to assessment” language. Consumers learn whether they qualify only after submitting a full application — at which point a hard credit pull is recorded against their bureau record and a decline becomes a permanent friction.

The data on this friction is severe:

• TransUnion Q4 2025 reports a 25%+ vehicle finance application rejection rate.
• WesBank received 78,983 new + 136,067 used finance applications in January 2026 alone.
• The Eighty20 Q4 2025 Credit Stress Report found a R3.7B single-quarter increase in vehicle finance overdue balances.
• Loan terms are creeping above 90 months as affordability tightens.

The result: a market in which the buyer cannot judge their own eligibility, the bank cannot price their risk without a hard pull, and the dealer cannot accurately forecast which leads will close. Every actor is operating with bad information.

2. The legal opening

A common assumption in the South African fintech community is that any tool offering an opinion on vehicle finance approval must be registered as either a credit provider (under the NCA) or a financial services provider (under the FAIS Act). This is incorrect.

• NCA Section 40(1) defines a credit provider by the act of extending credit. A simulator that does not extend credit is not a credit provider and does not require NCR registration.
• NCA Section 70 defines a credit bureau by the act of holding consumer credit information. A simulator that uses only self-reported data and never queries a registered bureau is not a credit bureau.
• FAIS Act Section 1 defines “financial product” as investments, insurance, deposits, and similar instruments. Credit agreements are explicitly excluded. A car finance estimator is therefore not regulated by FAIS and does not require an FSP licence.

What remains is POPIA, which governs the storage of any user-supplied data. Visio Approve handles this with explicit consent gates, an Information Officer registration, and a 90-day retention rule for any persisted simulation.

Inside this gap, Visio Approve operates as a lead generator + comparison tool, structurally identical to ClearScore SA, JustMoney, and RateCompare — but vehicle-specific and dealer-page native.

3. The affordability engine

The calculation has four ingredients:

1. Reg 23A minimum expense norms table. Published by National Treasury, it defines the minimum monthly living expenses a credit provider must assume for a borrower at any given gross income band. Encoded as a constant in lib/engine/affordability.ts.
2. Statutory deductions. PAYE + UIF approximation (banded by income).
3. User-supplied obligations. Rent, total monthly debt repayments, dependents.
4. Self-selected credit health. Excellent / Good / Fair / Poor / Don't Know — five buckets, mapped to the published rate margins of each of the four banks.

From these four inputs we compute:

• Net income = gross − statutory − Reg 23A min living
• Discretionary income = net − rent − other obligations
• Maximum sustainable instalment = discretionary × 0.65 (conservative buffer below the bank's own 80%-of-discretionary cap)
• Debt-to-income ratio at the proposed instalment
• Loan-to-value ratio for the specific vehicle and deposit
• Approval likelihood = weighted combination mapped onto Green / Amber / Red

Crucially, the rates returned to the consumer are not invented. They are pulled from a lib/banks/rates.ts constant containing the actual published rate cards of WesBank, Standard Bank VAF, Nedbank MFC and Absa VAF, indexed by SARB prime (10.25% as of January 2026). Each entry is timestamped with a verifiedAtdate and a public source URL — no fabricated numbers.

4. Why this matters for the buyer

A buyer using Visio Approve learns three things in 90 seconds that they would otherwise need three weeks and multiple branch visits to discover:

1. Whether they would probably be approved at this car's price point — not as advice, but as an estimate based on the same Reg 23A formula the banks themselves use.
2. What four banks would charge them at their self-reported credit band — and which is cheapest for their specific risk profile.
3. What the maximum vehicle price they can afford actually is — and how much extra deposit would flip an Amber estimate to Green.

This eliminates the most common failure mode in the SA used-car funnel: a buyer falls in love with a car that is structurally beyond their affordability, applies anyway, gets rejected, takes a hard credit hit, and disengages from the market for months. With Visio Approve, the same buyer is redirected to a vehicle they can afford before any bureau pull, and arrives at the dealership pre-qualified.

5. Why this matters for the dealer

The dealer F&I desk's central problem is wasted credit applications. Every rejected application costs roughly 30 minutes of an F&I manager's time, plus the opportunity cost of a lead that fails to convert. At a 25% rejection rate, a dealer doing 40 deals per month is wasting roughly 5 hours per week on applications that will never close.

Visio Approve embedded on a dealer's listing page filters this out. By the time a lead arrives at the F&I desk, they have already self-disclosed their income and obligations and seen the four bank rates side by side. The dealer can prioritise Green leads, reroute Amber leads to a deposit-coaching workflow, and decline Red leads with a simple “let's revisit when your DTI improves” — without ever burning a hard credit pull.

6. Revenue model

1. Bank lead-fees. Each of the four major lenders pays R150–R500 per qualified, pre-approved application. At 12,000 monthly completions × 30% handoff = ~R36M ARR.
2. Dealer subscription. R3,000/month for the integration, lead capture, and analytics. At 500 dealers = R18M ARR.

Combined steady-state: R54M ARR, 90%+ gross margin.

7. Limitations and ethical considerations

Visio Approve is an estimator, not a credit decision. Every screen carries: “This is an educational estimate. It is not a credit application, not financial advice, and does not guarantee approval. Final approval is subject to a credit assessment by the lender.”

We deliberately use a conservative buffer (65% of discretionary income vs the 80% banks themselves accept) to bias toward false negatives rather than false positives. A buyer told Amber who is then approved is delighted; a buyer told Green who is then declined is angry and loses trust. The asymmetry is reflected in the engine.

We do not store any bureau data and we do not call any bureau API. The user's self-reported credit band is treated as a hint, not a fact, and the engine surfaces the uncertainty: an Amber result is honest about why it is Amber.

8. Conclusion

Visio Approve is a small product with a precise legal positioning and a large market. By exposing the published Reg 23A formula, the four major banks' actual rate cards, and a conservative approval likelihood score to the consumer in 90 seconds, it removes the single largest information asymmetry in the South African vehicle finance market. It does this without becoming a credit provider, without becoming a credit bureau, and without crossing into FAIS-regulated advice — which is to say, without any of the regulatory burden that has prevented anyone else from building it.

The R54M ARR opportunity is not the most interesting thing. The most interesting thing is that it converts a trust deficit in the SA used-car market into a distribution channelfor the four major lenders — and that VisioCorp owns the layer that sits between them.

References: National Credit Act 34 of 2005 (Sections 40, 70, 81); Affordability Assessment Regulations 2015 (Reg 23A); FAIS Act 37 of 2002 Section 1; TransUnion Q4 2025 Mobility Insights; WesBank Vehicle Finance Barometer Jan 2026; Eighty20 Q4 2025 Credit Stress Report; SARB Repo Rate Jan 2026; WesBank, Standard Bank VAF, Nedbank MFC, Absa VAF published rate cards (verified 2026-01-01). Repository: github.com/Iamhamptom/visio-approve, commit 45c02d3. Live deploy: visio-approve.vercel.app.

Cite as: Hampton, D. (2026). “Visio Approve: A Neutral Multi-Bank Vehicle Finance Pre-Approval Layer for South Africa”. Visio Research Labs, VRL-AUTO-004.