The South African Reserve Bank wants to transform more than 30,000 ATMs into a shared network. Behind the proposal lies a bigger question: why should competing banks maintain separate infrastructure to deliver the same essential service?
South Africa is challenging one of the assumptions that has shaped modern banking: that the transition to digital payments should inevitably lead to the decline of cash infrastructure.
The South African Reserve Bank (SARB) is pursuing a different approach. Rather than trying to eliminate physical money, it wants to make cash cheaper to distribute, easier to access and less expensive to use. Its proposed reform would transform the country's fragmented, bank-branded ATM network into a shared, largely unbranded system supported by a national cash utility.
The initiative, with a pilot planned for January 2027 and an initial rollout of white-label ATMs expected before the end of that year, could fundamentally change how cash moves through South Africa's economy.
The proposal is significant because it challenges the traditional organisation of financial infrastructure. Instead of requiring every bank to maintain its own network of machines, cash-processing arrangements and distribution systems, the Reserve Bank wants to coordinate infrastructure that performs essentially the same function.
The question is not whether South Africa will eventually become more digital. It is whether the country can modernise its financial system without making access to physical money more expensive for the people who still depend on it.
The economics of an increasingly expensive payment system
South Africa's cash economy remains substantial despite the expansion of electronic payments.
According to figures cited in the Reserve Bank's 2026 Cost of Cash Research Study, approximately R180.65 billion was in circulation, while cash accounted for around 56% of consumer transactions by volume. The broader annual economic cost associated with cash was estimated at approximately R88.5 billion.
These figures describe different aspects of the economy. The amount of cash in circulation is a stock of money, whereas the annual cost estimate captures the broader resources consumed in distributing, accessing and using physical money.
The distinction matters. The R88.5 billion estimate is not simply the operating expenditure of ATMs, nor does it represent the amount the Reserve Bank expects its proposed reform to save. It reflects a much wider economic burden that includes cash handling, travel, time, fees and other associated costs.
Nevertheless, the figures expose a structural problem.
Cash remains important to consumers, retailers and informal businesses, but the infrastructure required to support it is expensive. Banks must maintain machines, secure cash, operate or contract cash-processing centres, and arrange regular replenishment.
These costs do not disappear simply because some consumers have switched to mobile banking.
An ATM that processes fewer transactions may become more expensive to operate on a per-transaction basis. Security, maintenance, connectivity and equipment costs continue even when usage declines.
Banks consequently face a difficult commercial decision. They can maintain machines that no longer generate sufficient activity to justify their operating costs, or they can close them and risk making cash less accessible to customers.
This is where a fragmented banking model becomes particularly inefficient.
When competing institutions maintain overlapping infrastructure, the costs of providing access to money may be higher than necessary. The problem is not that every ATM is redundant. It is that the system may duplicate facilities, logistics and operating arrangements that could be shared.
Why the Reserve Bank wants to change the rules
Under the traditional model, banks operate their own branded ATM networks or enter separate arrangements with independent operators. Their networks form part of their customer-service infrastructure and, in some cases, their competitive strategy.
The Reserve Bank's proposed national cash utility would change that arrangement.
Instead of treating each bank's ATM network as a separate system, the model would allow infrastructure to be shared across participating institutions. Customers could use common machines regardless of which bank holds their account, subject to the services supported by each machine.
The proposal also envisages integrating elements of the wholesale cash infrastructure, including cash centres and distribution arrangements.
This matters because the cost of cash is not generated by ATMs alone. Cash must be transported, counted, stored, processed and replenished. Improving the efficiency of the wider supply chain could deliver savings that simply removing bank logos would not achieve.
Consider the logistics of servicing multiple ATM networks in the same commercial area. If separate operators organise overlapping cash deliveries, there may be opportunities to coordinate routes, improve vehicle utilisation and reduce unnecessary movements.
The savings would depend on actual operating conditions, security requirements and cash demand. Shared infrastructure is not automatically cheaper in every location. But it creates an opportunity to optimise the system collectively instead of requiring every institution to solve the same logistical problems independently.
The underlying principle is straightforward: banks should be able to compete for customers without necessarily duplicating every physical component of the financial system.
What a shared ATM network could mean for consumers
The most immediate potential benefit is cheaper access to cash.
The Reserve Bank wants qualifying low-value transactions to become free below a threshold that must be established through the final arrangements. The intention is to make basic cash services less dependent on which bank owns the nearest machine.
For consumers, that could reduce the inconvenience of searching for a particular bank's ATM or paying additional charges when using another institution's machine.
The benefits could be particularly important for lower-income households, which have less room in their budgets to absorb transaction charges and travel expenses.
A relatively small ATM fee may appear insignificant to a wealthy customer. For someone managing a limited monthly income, repeated charges can become a meaningful expense.
The problem extends beyond transaction fees. If a nearby machine closes, the cost of accessing cash may include transportation, time away from work and the inconvenience of travelling to another location.
A shared network could help preserve access points that might otherwise become commercially unattractive to individual banks.
However, free withdrawals would not mean that cash distribution becomes costless. Machines still need to be purchased, maintained and secured. Cash must still be transported and processed.
The critical question is how those expenses will be funded.
If participating banks pay for the shared infrastructure, they may seek to recover their costs through other products or fees. If the utility charges institutions for services, the level and structure of those charges will influence whether consumers ultimately benefit.
The reform will be successful only if it reduces unnecessary costs and passes a meaningful share of the resulting savings to businesses and households.
Rural communities will provide the real test
The most important measure of the reform may be its impact on communities that banks have traditionally found less profitable to serve.
In major commercial centres, consumers often have several ATMs within a short distance. In rural communities and underserved urban neighbourhoods, a single machine may be an essential link between a bank account and the local economy.
Closing that machine can create significant hardship even when the decision makes commercial sense for its operator.
This exposes a fundamental difference between private profitability and public value.
An ATM may generate insufficient revenue to justify independent operation while remaining essential to a community. A shared utility could assess whether maintaining a single facility is more economical than requiring multiple banks to operate separate machines, while also considering the social cost of losing access altogether.
But consolidation must not become a justification for indiscriminate closures.
The number of machines is not the only relevant indicator. Their geographic distribution, reliability, cash availability and accessibility to low-income households are equally important.
A smaller network with reliable machines in the right locations could serve consumers better than a larger network concentrated in profitable districts. Conversely, a system that reduces operating expenses by removing essential rural facilities would undermine the financial inclusion it claims to support.
The Reserve Bank must therefore establish measurable service standards, including expectations for machine uptime, cash availability and access in underserved areas.
Without enforceable commitments, the promise of a national utility could remain an efficiency programme that benefits operators more than consumers.
The banks have reasons to resist
The proposal challenges more than the technical organisation of ATM infrastructure. It challenges established commercial interests.
Banks have invested in their networks, negotiated contracts with service providers and built customer relationships around their services. Their machines also provide visible branding and reinforce the convenience of remaining within a particular banking ecosystem.
A shared network could weaken some of these advantages.
Banks would retain the ability to compete over lending, savings, digital services, account fees and customer experience. However, ownership of the most convenient cash machine would become a less important differentiator.
The transition also raises difficult questions about existing assets and contracts.
Who will own or manage the machines transferred into the shared network? How will existing infrastructure be valued? How will costs be divided among participating banks? What happens to independent ATM operators and cash-processing companies? Who assumes responsibility for maintenance, security and service failures?
These questions will influence whether the proposed system creates genuine efficiencies or merely transfers costs and commercial risks from one institution to another.
There is also a competition concern. A shared utility could improve access, but if its pricing and operating decisions are not transparent, it could become a new bottleneck for the entire banking sector.
Smaller institutions must have fair access to the network, and consumers must have clear channels for resolving disputes.
The Reserve Bank's challenge is to establish a framework in which cooperation over infrastructure does not become an excuse for limiting competition elsewhere.
Security must improve alongside efficiency
Cash distribution presents security risks at every stage, from storage and processing to transportation and ATM replenishment.
A coordinated network could reduce unnecessary movements, improve route planning and create better visibility across cash operations. These changes could reduce some risks while improving the efficiency of cash delivery.
But centralisation also creates new vulnerabilities.
If too much of the system depends on one utility, a major operational disruption or technology failure could affect a large number of institutions simultaneously.
The system will therefore need contingency arrangements, reliable monitoring, clear accountability and sufficient operational redundancy.
Consumer protection is equally important.
When a transaction fails at an unbranded ATM, customers need to know which institution is responsible for resolving the problem. If an account is debited without cash being dispensed, or a customer experiences fraud, the shared model must not leave them navigating an unclear chain of responsibility.
The utility should establish consistent procedures for complaints, disputed transactions, service outages and suspected fraud.
A common infrastructure system can make access more convenient, but it must also make accountability clearer.
Cash and digital payments can coexist
The proposal challenges the idea that financial modernisation requires the rapid elimination of physical money.
Digital payments offer substantial benefits, including speed, convenience and the potential to reduce the need to transport and handle banknotes. But their effectiveness depends on reliable connectivity, electricity, suitable devices and affordable services.
Cash continues to offer practical advantages in transactions where electronic payment infrastructure is unavailable, inconvenient or unsuitable.
For informal traders, small businesses and households managing everyday expenses, physical money can remain an important part of economic activity even when digital payments are available.
South Africa's position is therefore pragmatic. Rather than forcing consumers to abandon cash, the Reserve Bank wants to make the infrastructure supporting it more efficient.
That approach recognises an important distinction: reducing the cost of a payment system is not the same as eliminating the payment method it supports.
The same principle has implications for other African economies. Countries can expand mobile payments and electronic transfers while still investing in reliable, affordable cash access. The two objectives are not inherently contradictory.
What Nigeria and the rest of Africa should consider
South Africa's proposed reform offers a useful case study for African policymakers facing the cost of financial infrastructure.
Nigeria, for example, has expanded digital payments and agent banking, but many small businesses and households continue to rely on cash for everyday transactions.
The policy question is not simply how to persuade consumers to use electronic payments. It is also how to ensure that the infrastructure needed to access and deposit physical money remains affordable.
Three lessons deserve attention.
First, regulators should examine whether competing institutions are duplicating expensive infrastructure unnecessarily. Shared facilities may offer opportunities to reduce costs, provided that they are economically justified and properly governed.
Second, financial inclusion must be measured by practical access, not merely by the number of bank accounts or digital payment products available. Consumers need to be able to use their money conveniently, whether through electronic transfers or cash.
Third, governments and regulators should avoid treating digital payments as a complete substitute for physical money before the infrastructure and economic conditions exist to support that transition.
However, South Africa's model should not simply be copied without examining local conditions. Differences in geography, security, banking structures, transaction volumes and regulatory capacity will influence whether a shared cash utility is viable elsewhere.
The broader lesson is that financial infrastructure should be designed around the needs of the economy rather than the historical boundaries of individual institutions.
The implementation will determine the outcome
The Reserve Bank's timetable envisages a pilot beginning in January 2027, the establishment of the national cash utility by March, the integration of cash centres during the middle of the year and an initial rollout of white-label ATMs before the end of 2027.
Those milestones are targets, not guarantees.
The transition will require agreement on governance, funding, ownership, pricing, operational responsibilities and the treatment of existing contracts. It must also demonstrate that a shared system can maintain service quality during the transition.
The most important indicators will be the actual cost of cash access, the reliability of machines, the availability of deposits and withdrawals, and the distribution of services across different communities.
Policymakers should also monitor whether savings are passed on to customers or absorbed elsewhere in the financial system.
The success of the programme should not be measured simply by how many machines lose their bank branding. It should be measured by whether consumers can access their money more cheaply, reliably and conveniently.
Conclusion: reforming the infrastructure, not abolishing the money
South Africa's proposed ATM overhaul is an attempt to solve a structural problem in the economics of banking.
Cash remains important, but the infrastructure supporting it is expensive. Individual banks have incentives to reduce their own operating costs, yet closing machines can leave communities with fewer convenient ways to access money.
A national cash utility offers a potential middle ground. It could reduce duplication, coordinate logistics and make basic cash services more accessible without requiring every bank to maintain a separate network.
The risks are equally real. Poor governance could create a new infrastructure monopoly. Cost-cutting could weaken rural access. Savings could fail to reach consumers, while unclear responsibilities could leave customers exposed when transactions go wrong.
The outcome will depend on whether the system combines operational efficiency with transparent pricing, fair competition, strong consumer protection and enforceable access standards.
South Africa is not trying to make cash disappear. It is questioning why the infrastructure needed to use cash should remain unnecessarily fragmented.
That is the more important idea behind the reform: financial modernisation does not always require replacing an existing service. Sometimes it requires redesigning the infrastructure that makes the service possible.
If the Reserve Bank succeeds, South Africa could demonstrate that a country can embrace digital payments while preserving affordable access to physical money — and that sharing essential infrastructure can sometimes deliver more public value than forcing every competitor to build its own.
The real measure of progress will not be whether cash survives. It will be whether ordinary people can access it without paying more than necessary.
Comments
Post a Comment