Reference🌍 Payments in Africa & the Middle EastIntermediate⏱ 26 min read

🇿🇦 Payments in South Africa

PayShap and its late takeoff, SAMOS and PayInc clearing, the debit order and its abuse, interchange set by the central bank, cash that costs ZAR 90 billion a year, and 96% inclusion that does not translate into usage

The framework: a central bank that operates as much as it regulates

The South African Reserve Bank (SARB) is South Africa’s central bank. It writes the payment system rules, runs the domestic RTGS, operates the regional RTGS for 16 countries, and, since 2025, owns half of the retail clearing house. Few central banks combine all four roles. A company looking for the relevant authority on any South African payment acceptance question therefore always ends up at the same institution, whether the issue is licensing, access to settlement, or a challenge to a clearing rule.

The national payment system rests on two statutes. Section 10(1)(c)(i) of the South African Reserve Bank Act 90 of 1989 tasks the central bank with establishing and conducting clearing and settlement systems. The National Payment System Act 78 of 1998, published in Government Gazette 19402 on October 28, 1998, gives the central bank responsibility for managing, administering, regulating, and supervising the national payment system, or NPS (SARB, NPS Vision 2030+ Consultation Paper, February 2026). These two laws alone establish the SARB’s authority over payments, and every South African payment requirement traces back to them.

RailOperatorSinceWhat it carries
SAMOS (South African Multiple Option Settlement)South African Reserve Bank–The domestic RTGS. Everything settles here in central bank money: real-time high-value payments, and the net positions submitted by the clearing houses
PayInc (formerly BankservAfrica)SARB (50%) / commercial banks (50%)1972Retail clearing: EFT credits and debits, cards, RTC, PayShap, DebiCheck. Designated the National Payments Utility by the SARB
RTC (Real-Time Clearing)PayInc2006The original near-instant credit transfer. It has stayed expensive, so in practice it is used for large amounts
PayShapPayInc2023The ISO 20022 instant credit transfer, with alias-based addressing (ShapID). The retail rail the whole country is betting on
DebiCheck and Registered MandatePayInc, under Payments Association of South Africa rulesRM since May 12, 2025Debit orders with an authenticated mandate, plus a fallback flow with a registered mandate
SADC-RTGS (formerly SIRESS)South African Reserve Bank2013Regional gross settlement, denominated in rand, for 16 southern African countries
TCIB (Transactions Cleared on an Immediate Basis)PayInc2021The instant retail counterpart to SADC-RTGS: low-value cross-border push credits
South Africa’s payment rails and their operators
🔑
BankservAfrica is now PayInc, and the central bank owns half of it
PayInc is South Africa’s retail clearing house, founded in 1972 and renamed PayInc in August 2025. At the same time it became a joint venture owned equally by the SARB (50%) and the commercial banks (50%). The Competition Commission approved the deal in 2025. The SARB’s modernization program designates it as the National Payments Utility, responsible for daily clearing and for owning and managing the schemes (SARB, Payments Ecosystem Modernisation, Industry Dialogue, 2026). Any document written before the rename calls it by its former name, BankservAfrica. To identify the operator referred to in a South African document, you therefore need to know when it was published.
ZAR 153.7T
settled in SAMOS in 2024, up from ZAR 150.1 trillion in 2023
SARB, National Payment System Regulatory and Oversight Report, 2024/25
16.0M
transactions settled in SAMOS in 2024, up from 13.9 million in 2023
SARB, National Payment System Regulatory and Oversight Report, 2024/25
88 % / 12 %
split of SAMOS value between real-time high-value payments and retail obligations submitted by the clearing houses
SARB, National Payment System Regulatory and Oversight Report, 2024/25

Self-regulation is the third layer of the framework, alongside the law and the central bank, and it is being dismantled. The Payments Association of South Africa (PASA), recognized by the SARB as a payment system management body (PSMB), sets the clearing house rules, admits participants, and imposes sanctions. The upcoming NPS Bill abolishes the PSMB model. The SARB has set up a PSMB Transitional Committee to reallocate PASA’s functions. They will go to the central bank itself, clearing house operators, an industry body, and other organizations (SARB, National Payment System Regulatory and Oversight Report, 2024/25). A contract that names PASA as the rule-making authority may therefore no longer point to the right party once those functions have moved.

PayShap: three flat years, then the breakout

PayShap is South Africa’s instant credit transfer system. Launched in March 2023, it is built on the ISO 20022 standard and operated by PayInc. It uses the ShapID for addressing, an alias built from the payee’s phone number and bank, in the form 072xxxxxxx@bank. An account number still works as an identifier. Funds reach the payee within seconds, at any hour of the day or night.

Adoption stayed low for three years because of pricing. Banks charged for every PayShap transaction, while EFT was free and cash was everywhere. A paid instant rail was competing with two payment methods that cost users nothing they could see. The SARB listed the under-adoption of the fast payment system among the problems its modernization program must fix (SARB, Payments Ecosystem Modernisation, 2026). Volumes jumped by an order of magnitude in the first five months of 2026.

905M
cumulative PayShap transactions at the end of May 2026, up from 461 million at the end of December 2025
ClearingPost / PayInc, 2026
≈ 89M/month
monthly run rate over the first five months of 2026, vs. ≈ 14M a month over the previous 33 months
ClearingPost / PayInc, 2026 (calculated from published cumulative totals)
ZAR 403B
in cumulative value at the end of December 2025, for an average transaction of about ZAR 874
ClearingPost / PayInc, 2026
6M
registered users, across 12 participating banks
ClearingPost / PayInc, 2026
March 2023
Launch
PayShap launches with the big banks. The initial limit is ZAR 3,000 per transaction, designed for small payments.
March 2024
PayShap Request and fraud prevention
Request-to-Pay goes live. The payee requests the payment, and the payer approves it in their app. A fraud intelligence system launches the same year (SARB, NPS Regulatory and Oversight Report, 2024/25).
August 2024
The limit rises to ZAR 50,000
The higher interbank limit takes PayShap beyond P2P payments alone (BankservAfrica, 2024; Absa, PayShap page, 2026).
March 2025
Ten banks, 251 million transactions
The SARB counts 4.6 million ShapIDs, more than 200 million payments, and more than ZAR 175 billion settled since launch, and estimates cumulative volume at 251 million transactions by the end of March 2025. Ten banks participate, and four more are expected within the year.
June 2026
One billion transactions
PayShap passes 1 billion cumulative transactions and tops 60 million a month (PayInc, June 2026).
2026
Merchant pivot
After three years focused on P2P, the operator formally repositions PayShap toward merchant payments and e-commerce, acknowledging that bank pricing and user experience held adoption back (PayInc, 2026).
⚠️
Pricing varies from bank to bank, and that is the problem
The PayShap scheme sets no prices. Each bank decides what it charges its customers per transaction. GoTyme Bank charges nothing at any amount, while the other large banks charge between ZAR 1 and ZAR 10 per transaction. Discovery Bank charges up to 0.5% of the amount, capped at ZAR 35, and Standard Bank has cut its price to a flat ZAR 2 up to the ZAR 3,000 limit (TechCabal, July 2026; TechCentral, 2026). The cost of a merchant payment therefore depends on the buyer’s bank, not the merchant’s. For the same cart at the same merchant, one customer may pay nothing and another several rand. The merchant can neither show that cost in its checkout nor standardize it, so its conversion rate depends on a variable it does not control.
Accepting a merchant payment through PayShap Request
Merchant
Sends a payment request
Through its PSP: amount, reference, and the payer’s ShapID or account
PayInc
Routes the request to the payer’s bank
Alias resolution, ISO 20022 message, scheme checks
Payer
Approves in their banking app
The payer sees the payee and the amount, then accepts or declines. A decline is a normal outcome that the merchant must handle
Payer’s bank
Debits and sends the push credit
Funds check, fraud checks, then irrevocable submission
Payee’s bank
Credits within seconds
Funds are available immediately; interbank settlement follows in SAMOS
ℹ️
A push credit cannot be recalled
PayShap is an irrevocable push payment, with no chargeback mechanism comparable to cards. A refund requires a reverse transaction initiated by the party that received the funds. A social engineering victim who approved the payment in their own app has no recourse through the rail. As far as the system is concerned, the transaction is valid. Prevention has to happen before approval, through a clear payee name shown to the payer and through the sending bank’s fraud controls.

RTC and EFT: the credit stack PayShap did not replace

South Africa has three credit transfer rails. PayShap was added to the two existing systems without replacing either. EFT credit is the original bulk transfer: cleared in batches, cheap, and slow. RTC has offered near-instant credit since 2006, but its price has confined it to large amounts, while PayShap covers small instant payments. EFT credit still carries by far the most value of the three.

1.13B
EFT credit transactions in 2024, up from 406.9 million in 2010
SARB, National Payment System Regulatory and Oversight Report, 2024/25
≈ ZAR 12.9T
in EFT credit value as of March 2025, up from ZAR 4.82 trillion in 2010
SARB, National Payment System Regulatory and Oversight Report, 2024/25
90 %
of retail payment value carried by EFT credit in 2022
SARB, Digital Payments Roadmap, April 2024 (PAYIR data)
6.4B
retail transactions in 2022, up from 5.4 billion in 2021 (excluding cash withdrawals)
SARB, Digital Payments Roadmap, April 2024 (PAYIR data)
CriterionEFT creditRTCPayShap
SpeedBatch clearing, delayed creditNear-instantA few seconds, 24/7
Cost to the senderLowestHigh, its long-standing weaknessFree to ZAR 10, depending on the payer’s bank
Typical amountPayroll, suppliers, high volumesLarge amounts, urgent paymentsSmall amounts; ZAR 50,000 interbank limit
AddressingAccount number and branch codeAccount numberShapID (alias) or account number
Message standardLegacy batch formatsLegacy formatsISO 20022
Merchant useRare for collections, common for payoutsMarginalStated target since 2026
Choosing a credit transfer rail in South Africa

Checks were withdrawn from the South African payment system effective December 31, 2020. The decision was made jointly by the SARB, the Financial Sector Conduct Authority, PASA, and the Banking Association South Africa (SARB, Digital Payments Roadmap, April 2024). The reasons given were processing times, fraud, cost, limited acceptance, and declining use. South Africa thus scrapped an instrument that several developed markets still keep in circulation.

⚠️
Instant EFT: a workaround, not a rail
“Instant EFT” is an online payment method in which the customer is redirected to their online banking and sends the transfer themselves. The provider then confirms that the transfer was made. A large share of South African e-commerce runs this way. Ozow is the best-known provider, with more than 6 million registered users (Ozow, 2026). It is registered with PASA as a System Operator and a Third Party Payment Provider, so Ozow is neither a bank nor a scheme. Some of these services relied on collecting the payer’s online banking credentials, known as screen scraping, which the SARB has regulated under Directive 2 of 2024.

The debit order, its abuse, and the DebiCheck fix

The debit order is South Africa’s direct debit: a creditor instructs the payer’s bank to debit the payer’s account. It is used for insurance, consumer credit, subscriptions, and school fees. For years it also fueled an industry of abusive collections. At the time, mandates were not registered anywhere at the payer’s bank. The bank had no reference to check that a collection matched a real commitment. A collector could submit whatever amounts and frequencies it chose without producing an enforceable mandate.

A race to be first in the queue grew on top of this setup. Early debit orders were submitted ahead of every other flow, starting on payday, to grab the balance first. Disputes rose sharply. Banks refunded, collectors resubmitted in the next cycle, and payers lost control over the order in which their accounts were debited. DebiCheck was introduced in response: the payer authenticates the mandate with their own bank when signing the contract. Without an authenticated mandate, the payer can still dispute the collection, and the creditor has no proof of agreement to counter the dispute.

The life cycle of a DebiCheck mandate
Creditor
Initiates the mandate
Maximum amount, frequency, first collection date, contract reference
PayInc
Routes the authentication request
To the payer’s bank, which presents it through its own channel (app, USSD, card at a POS terminal)
Payer
Authenticates or lets it expire
No response is where the system leaks: 38% of mandates did not go through in 2024
Payer’s bank
Registers the mandate
Later collections are checked against the registered mandate: amount, frequency, creditor
Creditor
Collects in the morning window
A collection that matches the authenticated mandate is hard to dispute, which is the whole point of the flow
58.8M
DebiCheck mandates initiated in 2024 (43.6M in 2022, 52.4M in 2023)
SARB, National Payment System Regulatory and Oversight Report, 2024/25
62 %
average DebiCheck mandate authentication rate in 2024 (59% in 2022, 60% in 2023)
SARB, National Payment System Regulatory and Oversight Report, 2024/25
200.1M
DebiCheck and RMS collections processed in 2024, up from 167.8 million in 2022
SARB, National Payment System Regulatory and Oversight Report, 2024/25
79 %
collection success rate, stable from 2022 to 2024
SARB, National Payment System Regulatory and Oversight Report, 2024/25
⚠️
62% authentication: the number that decides the business model
DebiCheck’s technical success rate is 98%, but its authentication rate is 62% (SARB, 2024/25). The gap comes from payers not responding to the authentication request, not from message processing failures. Nearly four in ten mandates therefore fail on the first attempt. A biller that relies on DebiCheck to sign up customers builds that loss into its forecasts, follows up through another channel, and tracks the authentication rate channel by channel. The industry is working to raise the figure, which is climbing two points a year.
FlowMandateProcessing windowRisk to the creditor
DebiCheckAuthenticated by the payer with their bankFirst morning windowLowest: each collection is checked against the registered mandate
Registered Mandate (RM)Registered, not authenticated by the payerEveningMedium: the mandate is on file, but the payer has not actively confirmed it
Legacy EFT debitNo mandate registered with the bankLate eveningHighest: the bank must refund without examining whether the mandate is valid
The three debit order flows, and why processing order matters

The Registered Mandate Service (RMS) was a transitional arrangement for creditors whose payers had not responded to the DebiCheck request. RMS initiations fell 5% in 2024 to 9.5 million, precisely because DebiCheck authentication had improved (SARB, 2024/25). The Registered Mandate (RM) replaced it on May 12, 2025 as a stand-alone payment system, moving from the morning window to an evening window. Nothing carried over automatically: a mandate signed under the old service does not migrate to the new one on its own.

ℹ️
Payroll deductions are the next regulatory front
A payroll deduction takes a debt out of the debtor’s pay, outside the payment system. Some creditors turn to it when the debit order fails. In July 2024, the SARB published a draft directive governing these payroll deductions. It sets out which deductions are permitted and under which payment instructions. The steering committee includes the SARB, National Treasury, the Department of Employment and Labour, the dtic, the National Credit Regulator, the DPSA, and the FSCA (SARB, NPS Regulatory and Oversight Report, 2024/25). The stated goal is to stop employees from being exploited through collections embedded in payroll.

Cards: the dominant rail, with costs set by the central bank

South Africa is Africa’s largest card market. Card purchases made up 65% of retail payment volume in 2022 (SARB, Digital Payments Roadmap, April 2024). There is no domestic scheme: Visa and Mastercard account for almost all issuing and acquiring. Only locally licensed firms can access the market, because Directive 1 of 2020 bars foreign issuers and acquirers not licensed under South African law from issuing or acquiring domestic cards.

South African interchange is set by the central bank itself, following the Interchange Determination Project launched in 2011. This differs from the European model, where the schemes set the rates and a regulation then caps them. The card phase concluded in 2014, used a cost-based methodology, and replaced the three rates then in force with 12 rates. Those rates have applied since January 1, 2015 (SARB, Interchange Determination Project, Phase 2, card results, 2014).

Transaction typeDebitCredit
Card present, EMV-compliant0,44 %1,48 %
Card not present, 3-D Secure-compliant0.43% to 0.53%1.57% to 1.89%
Cash withdrawal at the point of sale (cash-back at POS)ZAR 1.11 per transaction–
Interchange set by the SARB, effective January 1, 2015
0.36% to 2.45%
card interchange range, by transaction type and card type
SARB, Digital Payments Roadmap, April 2024
1.5% to 3.5%
card processing fee charged to merchants
SARB, Digital Payments Roadmap, April 2024
8 %
of micro, small, and medium-sized enterprises have a card terminal (*speed point*)
FinScope South Africa MSME Survey 2020, cited by the SARB, 2024
0,2 %
of consumers paid by QR code in 2022
FinScope, cited by the SARB, Digital Payments Roadmap, April 2024
⚠️
The cost of acceptance is what holds acceptance back
The SARB writes that high fees discourage small merchants from accepting digital payments because they eat directly into margins (Digital Payments Roadmap, April 2024). Terminal penetration shows the scale of the gap: more than 90% of merchants have no terminal. Cards dominate retail payment value in a country where in-person card acceptance is still largely confined to the formal economy. PayShap’s merchant pivot and the planned interoperable QR standard aim to extend acceptance beyond it.
📱
Two merchant QR networks, no interoperability
SnapScan (FireID Payments, backed by Standard Bank) claims more than 60,000 merchants, and Zapper (Wonderwill Ltd) 31,000 business customers (company figures, 2026). They are two competing closed loops: acceptance depends on the app, not on a national standard. The industry is working on an interoperable QR standard, which the SARB has written into its Roadmap.
🛡️
Fraud has moved online
Card fraud losses rose 26.2% in 2024 to ZAR 1.466 billion, even as overall financial crime fell 18% to ZAR 2.7 billion. Card-not-present fraud accounts for 85.6% of gross losses on South African-issued credit cards (SABRIC, Annual Crime Statistics 2024, published August 28, 2025).
🏪
A contested acquiring market
Acquiring is split between acquiring banks and specialist providers. Yoco, Peach Payments, PayFast, and Ozow are among the names merchants will come across. The competition is over small merchants, precisely where terminal penetration is lowest.
🧾
A mature, already regulated BNPL market
PayJustNow (Weaver Fintech), Payflex (2017), MoreTyme (GoTyme Bank), and Mobicred all operate in the market. Their legal classification needs a close look. PayJustNow offers an interest-bearing “Pay in 12” alongside its interest-free pay-in-three, and the two fall under different regimes under the National Credit Act.

Cash: massive, costly, and now officially accepted

Cash underpins the everyday economy for a large share of South Africans, including many who are paid electronically. In its Roadmap, the SARB puts it plainly: cash remains king. The documented pattern is the pass-through account: money comes in and goes straight back out.

48 %
of adults withdraw all their money as soon as it hits their account
FinScope 2022, cited by the SARB, Digital Payments Roadmap, April 2024
34 %
of banked adults are “mailbox” users: the account is used only for withdrawals
FinScope 2022, cited by the SARB, Digital Payments Roadmap, April 2024
97 % / 74 %
of MSMEs take payments in cash / pay their staff in cash
FinScope South Africa MSME Survey 2020, cited by the SARB, 2024
≈ ZAR 90B
annual cost of cash to the South African economy
SARB, Cost of Cash Study, November 2025

The annual cost of cash splits almost evenly into two categories. Direct costs (withdrawal and deposit fees) come to about ZAR 43.5 billion, or 49%. Indirect costs reach about ZAR 46.1 billion, covering travel and waiting time, crime and losses, opportunity cost, and costs passed on by merchants (SARB, Cost of Cash Study, November 2025). These costs fall hardest on low-income households. Cash is not free for its users. They pay in time, theft risk, and higher prices rather than in a line item on a bank statement.

🔑
The SARB changed course in June 2026
In its June 2026 position paper, “Towards a Cash Smart Society,” the SARB describes cash as a lasting part of a hybrid payments economy, to be actively managed rather than left to decline. The central bank is thus dropping the goal of replacing cash that shaped its earlier documents. The plan centers on a national cash utility jointly owned by banks, retailers, and other parties. It adds a framework for white-label ATMs, cash-back at the till, and universal service obligations to prevent cash deserts. A payment acceptance strategy for South Africa should therefore build in cash for the coming decade, not plan for its disappearance.

Withdrawal statistics point the other way. ATM transactions fell from 509.4 million in 2023 to 289.4 million in 2024 (SARB, NPS Regulatory and Oversight Report, 2024/25). The central bank sees this as a shift to digital. The ATM network is also being consolidated, which explains why white-label ATMs feature in the Cash Smart strategy. The decline is in the number of transactions, not in demand for cash: as access points thin out, people may take the same amount in one withdrawal instead of several.

ℹ️
M-Pesa failed in South Africa because too few people were unbanked
Vodacom launched M-Pesa in South Africa in 2010 and shut the service down in June 2016 with about 76,000 active users, against an initial target of 10 million (Vodacom, 2016). The Kenyan model targeted people without bank accounts, for whom a mobile wallet was the first way to store and send money. Most South Africans already had an account. The service offered them neither new access nor a wider acceptance network than cards. Any project that applies East African adoption assumptions to South Africa runs into this different starting point.

Inclusion: 96% have accounts, 19% use them weekly

Formal inclusion means holding at least one regulated financial product: insurance, payments, savings, or credit. South Africa’s formally included population rose from 91% in 2021 to 96% in 2022, and its banked population from 81% to 82% (FinScope, cited by the SARB, April 2024). Few emerging markets reach that level. But these rates measure ownership of a product, not its use, and in South Africa the two diverge sharply.

96 %
of adults are formally included (insurance, payments, savings, credit)
FinScope 2022, cited by the SARB, Digital Payments Roadmap, April 2024
82 %
of adults have at least one bank account
FinScope 2022, cited by the SARB, Digital Payments Roadmap, April 2024
> 70 %
of banked adults use their account once a month or less
FinScope, cited by the SARB, Digital Payments Roadmap, April 2024
19 %
of banked adults use their account every week
FinScope, cited by the SARB, Digital Payments Roadmap, April 2024
ReasonShare of respondentsWhat it means for an issuer
Not enough money50 %No payment product can fix an income problem
Preference for cash28 %The lever is merchant acceptance, not financial education
Account or card too expensive6 %Withdrawal and transaction fees are a documented barrier
Avoiding debit orders and standing orders6 %The legacy of abusive debit orders still undermines trust
Distrust of banks2 %Marginal, contrary to popular belief
Merchants don’t accept it0,1 %Reported as negligible; the perceived barrier is cost, not acceptance
Why accounts go unused: reasons given

Social grants play a major role in retail payment flows. South Africa’s grant system reaches more than 11 million people and benefits more than 40% of households. At the end of December 2021, the South African Social Security Agency (SASSA) had 11,550,605 beneficiaries. 63% (7,285,934) were paid through the South African Post Office and 37% (about 4,264,671) into a commercial bank account (SASSA, Annual Performance Plan 2022-23, cited by the SARB). Grant payment day brings the month’s peak in withdrawals, and that spike is visible across all retail flows.

ℹ️
South African retail banking was built from the bottom up
Capitec has more than 24 million active customers (February 2025), built by bringing low-cost accounts and cards to lower-income households. GoTyme Bank, formerly TymeBank, a licensed digital bank (FSP 49140, NCRCP 10774), has taken that approach further with its MoreTyme BNPL product, available at more than 3,000 partner retailers (GoTyme Bank, 2026). South African inclusion was built on deposit accounts and cards, delivered by low-cost retail banks. That sets it apart from the East African model, where mobile-operator wallets played that role.

Digital payments require mobile data, and its price is a real constraint in South Africa. In 2022, a gigabyte cost up to ZAR 85 in South Africa, compared with $1.53 in North Africa and $2.24 in Western Europe (SARB, Digital Payments Roadmap, April 2024). The Competition Commission had already found, in its Data Services Market Inquiry of December 2, 2019, that poor consumers paid inexplicably more per megabyte than wealthy ones. A payment app that eats into a data bundle becomes unusable as soon as the bundle runs out, which shuts out part of the very population it targets.

The regional picture: the rand as southern Africa’s settlement currency

SADC-RTGS, formerly SIRESS, is southern Africa’s gross settlement system, operated by the SARB since 2013 on behalf of the SADC Committee of Central Bank Governors. It settled in rand, and only in rand, until the Angolan kwanza joined on July 27, 2026. Sixteen countries participate: South Africa, Angola, Botswana, Comoros, the DRC, Eswatini, Lesotho, Madagascar, Malawi, Mauritius, Mozambique, Namibia, Seychelles, Tanzania, Zambia, and Zimbabwe (SARB, SADC-RTGS page, 2026). The rand therefore remains the de facto settlement currency for high-value payments in southern Africa, even for transactions in which neither party is South African.

  • 88 participants connected to SADC-RTGS as of March 31, 2025, including 9 South African and 14 Zimbabwean participants (SARB, NPS Regulatory and Oversight Report, 2024/25).
  • October 2024: Common Monetary Area flows (South Africa, Eswatini, Lesotho, Namibia) migrate to SADC-RTGS. The April 2024–March 2025 financial year is the best in the system’s history.
  • TCIB (Transactions Cleared on an Immediate Basis), operated by PayInc since 2021, is its retail counterpart: low-value cross-border push credits, cleared immediately and settled later.
⚠️
Low-value cross-border payments in the CMA broke, then got fixed
Small transfers between South Africa and its CMA neighbors used to run through the domestic retail system. That setup no longer let banks comply with FATF Recommendation 16, which covers the information that must accompany transfers, so it was dismantled. Banks had to move these flows in a hurry to SADC-RTGS, a high-value rail. The SARB has documented the result: slower transfers and higher costs for customers. In March 2025, the central bank finalized a directive requiring these flows to move to TCIB, the system designed for them (SARB, NPS Regulatory and Oversight Report, 2024/25).

Amendments to FATF Recommendation 16 went out for consultation in February 2025, followed by industry discussions in March 2025. Once adopted, they will require a revision of Directive 1 of 2022, which applies those requirements to the national payment system. The corridors linking South Africa to Zimbabwe, Malawi, and Mozambique therefore face changes to identification requirements soon. Informal flows in these corridors are huge, and TCIB was built to bring them into the formal system.

ℹ️
Khokha is not a retail CBDC
Project Khokha, launched in 2018 with the Intergovernmental Fintech Working Group, is a research program on wholesale CBDC, focused on securities tokenization and interbank settlement on a distributed ledger. The SARB has also taken part in mBridge-style cross-border projects. No decision has been made on a retail digital rand. The work done under the Khokha name is an experiment among financial institutions and involves no instrument that consumers or merchants can use.

Operating in South Africa: the regulatory calendar and what to do

South Africa has launched its biggest payments overhaul since 1998, with three workstreams moving in parallel. The NPS Bill rewrites the law, the PEM program (Payments Ecosystem Modernisation) delivers the technical and institutional modernization, and Vision 2030+ prepares the next strategic cycle. Their timelines overlap, so a project launched today will see the legal framework, the clearing infrastructure, and industry priorities change one after another.

2018
Vision 2025
The SARB publishes the National Payment System Framework and Strategy: Vision 2025, which sets nine industry goals, 26 strategies, and an action plan of 106 actions. Everything that follows stems from it.
April 18, 2024
Digital Payments Roadmap
A 17-action plan to tackle barriers to digital payment adoption: financial literacy, connectivity, power supply, acceptance costs, acquiring infrastructure.
May 17, 2024
Cybersecurity directive
Published in Government Gazette No. 50665, effective August 17, 2024. It requires payment institutions and market infrastructures to put a cyber resilience framework in place.
November 15, 2024
Directive 2 of 2024 on screen scraping
Government Gazette No. 51556. Companies that collect online banking credentials to initiate a transfer on the payer’s behalf must register with the SARB. An interim measure until an open banking framework is in place.
March 2025
Three consultations and a bill
Industry workshops on the NPS Bill, a consultation on interoperability (March 11), a consultation on cloud and offshore hosting, and publication of the exemption from the Banks Act 94 of 1990 that gives non-banks direct access.
August 2025
BankservAfrica becomes PayInc
The rename, and the SARB takes a 50% stake. The operator is designated the National Payments Utility.
February 2026
Vision 2030+ consultation
The SARB puts an overarching objective, “inclusive growth and development,” and 12 ecosystem trends out for consultation. Workshop on March 9, 2026; responses due by March 31, 2026.
June 2026
Towards a Cash Smart Society
The central bank commits to keeping cash and begins setting up a national cash utility.

Until now, direct access to the payment system has been reserved for institutions with a banking license. The NPS Bill, sponsored by National Treasury with SARB support, is meant to open it to non-banks. Until the bill becomes law, the central bank has prepared an exemption from the Banks Act 94 of 1990. It removes certain payment activities from the definition of the business of a bank and places them under a dedicated regulatory and supervisory framework. The exemption and its framework went out for consultation in March 2025 (SARB, NPS Regulatory and Oversight Report, 2024/25). Before the new law takes effect, this exemption is the fastest route to direct access for a non-bank provider.

🔑
PEM’s goals, and their impact on payment providers
The PEM program has three measurable objectives: support inclusion through the adoption of affordable digital payments, establish an efficient national utility, and open participation to non-banks. It publishes three illustrative targets for 2030. The first is a gradual shift from cash to digital. The second is access for 100% of the economically active population to at least two digital channels. The third is pricing parity, where the cost to participants does not exceed the actual cost of the service (SARB, PEM, Industry Dialogue, 2026; targets presented as illustrative). The SARB itself acknowledges that it does not control every lever. Hitting these targets depends on investment and pricing decisions by banks and providers.
  • Pin down your status first: bank, System Operator, Third Party Payment Provider, or user of the Banks Act exemption. That status, not the technology you use, determines your access to the clearing houses.
  • Never collect through legacy EFT debits. The payer is entitled to a refund there, with no review of the mandate. Move to DebiCheck, or failing that to the Registered Mandate.
  • Budget for a 38% authentication failure rate on DebiCheck mandates, and track the rate by channel.
  • Demand a fee breakdown: SARB-set interchange, scheme fees, acquirer margin. Published ranges run from 0.36% to 2.45% for interchange and from 1.5% to 3.5% for the total fee.
  • Build cash into the customer journey, for both collections and refunds. Nearly one adult in two empties their account as soon as money comes in.
  • Treat PayShap as an emerging merchant rail, not a proven one: the cost the payer sees depends on their bank, and push credits cannot be charged back.
  • Track the NPS Bill and the PSMB transition: PASA is not set to remain the rule-making body.