African Families, African Money
Bridging the Money Transfer Divide
A study on the South African money transfer
environment for FinMark Trus t
For many years South Africa’s urban areas have provided employment for migrants and
immigrants from other parts of South and Southern Africa. Money transfers from these migrants &
immigrants constitute an important source of income to their families and relatives in other parts
of the country and continent. This study considers the availability and features of the range of
money transfer services available in South Africa. In particular the study compares the costs and
benefits of the various products provided by formal providers to the informal money transfer
services often used by low income individuals. As money transfers are one of the most important
financial services required by low income individuals, the formal sector’s ability to provide
products that are competitive with respect to informal products is an important requirement for
The money transfer environment is divided in two ways:
§ Formal and informal services. Many individuals use informal money transfer
mechanisms – relying on a friend or taxi driver to act as courier to a rural recipient.
Formal services are divided into bank services, post office and specialist money agent
services (i.e.: Western Union & Money Gram).
§ Domestic and Cross border services. Transfers across borders face substantially
different technical, legal and political environments. Whereas governments are broadly
supportive of attempts to formalise domestic transfer services, the same cannot be said
for cross border transfers. Cross-border transfers, involve people who do not vote,
provide opportunity for cross border money laundering, could violate exchange controls
and might even encourage illegal migration.
Some ‘products’ are presented to the market as complete and independent money transfer
products (i.e.: postal orders) and other’s are offered only in conjunction with other services i.e.: a
bank account can be used for person to person (“P2P”) transfers but this is only part of a bundle
of services provided. The report ”unbundled” such services to understand only the costs and
features required to affect P2P transfers.
Domestic product / services
The Terms of Reference for this report did not include a formal investigation into the scale and
range of informal products used by different income groups, however, anecdotal evidence was
used to describe the two informal products that seem to predominate in South Africa: using
friends and taxi drivers as money carriers.
Postal products include postal orders, telegraphic money orders and the new PIN money order
which provides an on-line immediate funds transfer service to any post office.
Banks offer two basic transfer products: by electronic transfer, if both the sender and the recipient
have a bank account, and by direct cash deposit, if only the recipient has a bank account. Both
products require that both parties have access to banking facilities. More specifically, the recipient
must actually be banked and have access to the bank’s distribution network and the sender must
have his/her own bank account or have access to a branch of the recipients bank (although by
special arrangement some banks do accept cash deposits on behalf of clients of other banks, this
is not widespread practice). In this report banking services are categorised as:
§ Single recipient account : only the recipient has an account and the sender deposits into
a branch of the recipient’s bank. Whereas the banks offer this simple deposit account
product, interviews with frontline staff in several of the major banks indicated that in
reality they did not encourage customers who do not receive a regular payroll generated
§ Two accounts: both parties have accounts that are used exclusively to affect transfers.
§ Sender’s transaction account: The sender uses the account to process wages and the
recipient uses his account only to accept transfers. Strictly speaking, this makes it
incomparable to other money transfer products because it includes additional services to
transfers. It is included for sake of completeness and because it is probably more
realistic then the latter two.
The graph below compares the derived cost to the user of each of the products/services
identified. The shaded blocks shown against the informal channels depict the real, but un-
quantified risk of a loss of funds when using these channels.
Comparison of the cost to transfer R250 using different products/channels
Source: Genesis Analytics
Cost of R250 Transfer
Single Recepient Account
Sender's Transaction Account
Secure Postal Order
Telegraphic Money Order
Pin Money Order
Findings: Domestic market
§ The average cost (across the three bank products identified) for a monthly bank based
transaction account is probably in the order of R30, though this does not capture the
utility that the account holder derives from other banking services
§ The new Post Office PIN product, at R21, is cheaper then the average cost of a bank
based product. Moreover, funds are available within minutes across a network that
reaches many of the less developed parts of the country.
§ For amounts above R250, bank based transfer methods become increasingly
competitive, as the fees are fixed.
The product that scores most highly on general accessibility is the PIN money order. This new
product may seriously rival banking products, because:
§ no start up costs or monthly fees are required
§ of the rural and nationwide reach of the post office
§ of the products simplicity
§ of the instant availability of the funds to recipient
Bank offerings are relatively less attractive from the perspective of overall accessibility given the
poorer coverage of the banks in rural areas.
Overall it seems that the ability to transfer funds around the country is less of a constraint than is
sometimes thought. Utilising either the banks (recipient banked) or the post offices (both parties
unbanked) low-income individuals can make R250 transfers for between R10 and R30.
Compared to informal products, these are at least as attractive on price, present far lower risk,
and, with the new PIN money order, compete on accessibility.
In our analysis we have assumed the sole benefit to a recipient of having a bank account is to
receive transfers. It seems likely that with a bank account and a growing need to make payments
(for mobile recharge, utility payments etc) or receive state transfer payments, as well as to save,
more and more people will reap the benefits of having a bank account & would be able to make
payments at the very low cost associated with inter-account transactions. Thus providing a low
cost money transfer solution for the poor may be resolved by an overall drive to provide bank
accounts and possibly a requirement that banks accept deposits on behalf of other banks. The
Post Office PIN product does to some extent disintermediate the banks, and this could create a
problem for any market entrant that wishes to provide a money transfer service as part of a core
The report considers four categories of cross border transfer:
• Informal transfers – relying on friends and taxi drivers as couriers
• Post office products – money orders
• Bank products – the banks are able to effect P2P transfers between banked individuals &
execute bank drafts where the bank sends a cheque to the recipient
• Money transfer agents – the product/service provided by Money Gram and Western
All providers of formal money transfer products are required to be at least a limited authority
authorised dealer in foreign exchange (as defined in the Exchange Control Act) and to act,
comply and enforce an increasing array of regulations, ensuring:
• That, the appropriate tax has been paid on funds to be transmitted
• That the sender has the appropriate residency, immigration or work documents
authorising the sender to earn Rands
• That the sender is neither in breach of exchange controls nor has exceeded their limit for
the category of funds to be transmitted
• That the funds to be transmitted are the result of bone fide income generating activities
and are not the proceeds of crime
Faced with this task, most institutions either discourage transfers from low income individuals
whose bone fide may be more difficult to ascertain or, alternatively, have a strong incentive not to
comply with these regulations.
Formal cross border payments are also considerably more costly than their domestic
counterparts. The graph compares the price of the products identified in this study.
Cost of completing a R250 international transfer
Source: Genesis Analytics
Money Transfer Agents
Secure Postal Order
Telegraphic Money Order
Findings: International Market
• Formal cross border products are considerably more expensive than local transfers and
are prohibitively expensive for small amounts
• Given the cost of formal products it is no surprise that informal products remain important
for countries that border South Africa. In some instances informal products are the only
viable means of affecting a cross border transfer given the lower likelihood of the
recipient having a bank account in the receiving country. In the case of Zimbabwe the
huge gulf between the black market and official exchange rate impose a terrible penalty
on anybody that utilises formal products. Unfortunately informal mechanisms expose the
sender/recipients t`o considerable degrees of risk if the courier becomes the victim of
• Although cross border Post Office products are competitive priced they are not
competitive when it comes to speed of transfer and security. The quality and efficiency of
the post office in the receiving country may vary and there may or may not be links
between the South African Post Office and the post office in the receiving country.
• Bank products (P2P transfers and bank drafts) cost around R150 per transaction. This is
because the banks continue to charge a SWIFT fee and commission on each transaction,
even when funds are transferred to subsidiaries of the same bank in other countries, and
even if these countries are in the CMA, e.g.: Standard Bank of Lesotho or Swaziland
• The money transfer products of Western Union and Money Gram are cheaper than bank
products (R100 per transaction) and have the advantage of large networks in recipient
countries and that the funds are available immediately. Western Union is currently not
operating in South Africa. As Western Union provides one of the only means of transfer
for sending money to unbanked recipients the report investigated the circumstances
surrounding Western Unions withdrawal from South Africa in December 2001.
Western Union in South Africa
Western Union is the world largest money transfer company with a 24% market share and
150,000 locations worldwide. Migrant communities all around the world use Western Union to
affect money transfers to family and friends living in their country of origin.
In 1995, Western Union initiated operations in South Africa through Union African Money
Transfers. In the course of their operations, UAMT developed a network of retail outlets, which at
its peak numbered over 150 points of representation.
It seems that the burden of enforcing compliance with exchange controls in South Africa proved
too much for the UAMT management, and this did not go unnoticed by the Reserve Bank.
Matters came to a head with the implementation of new balance of payments reporting
requirements. In early 2001, the SARB reduced the time period that was allowed to lapse before
a financial transaction was reported to them from 1 week to within 24 hours. The cost of
implementing a system that could support this level of reporting made many of UAMT’s outlets
unprofitable with the result that UAMT reduced its network to 17 outlets. These outlets became
increasingly overburdened as the business of over 150 outlets converged on these 17 sites. As a
result, service quality and speed of service dramatically decreased.
In addition, although the exact nature of the regulator’s concerns with the Western Union
operation are confidential, there seems to be a reasonably widespread view that UAMT not only
struggled to implement appropriate systems, but were not overly committed to observing the spirit
and the letter of exchange controls. Matters were complicated further with the Reserve Bank’s
circular of October 2001 that indicated its desire to improve the enforcement of exchange controls
including a prohibition on the net settlement of foreign exchange transactions.
Net settling means that agents (UAMT) and Western Union only settle with one another after a
specified period, at which point only “net” balances are actually paid between agents and WU.
With net settlement, UAMT was able to use funds collected in SA for outward transfers to pay SA
recipients of inward transfers. With the enforcement of the net settlement prohibition, UAMT was
no longer able to pay money out to local recipients from money that was coming in from local
senders who were remitting overseas—they were made to wait until the actual funds destined for
local recipients had arrived. This compromised the Western Union product that guarantees the
immediate availability of funds to recipients. To sustain the relationship Western Union would
have had to increase its credit exposure to the agent. It is believed that the deterioration in
service quality and speed, the rising cost of exchange control compliance and poor execution by
UAMT, as well as increased and increasing credit exposure, led Western Union to end its
relationship with UAMT and suspend operations in South Africa.
Part of Western Unions problems seem to have been a structural mismatch in the business
model between an increasingly sophisticated exchange control compliance and reporting regime,
in which the Reserve Bank would prefer only banks to act as authorised dealers, and the
entrepreneur operator which was originally appointed by Western Union in South Africa.
The main challenge with respect to domestic transfers relates to network density and
complimentarily1. The banks provide safe and cost effective money transfers providing, that at
least the recipient is banked, and have good distribution in urban areas but not necessarily in the
rural areas where recipients are most likely to be located. If banks were to accept deposits on
behalf of each other this would allow those institutions that focus on rural distribution to compete
more effectively and would significantly increase network density from the perspective of the
sender who wishes to support a rural recipient, who at best, has access to a branch or ATM of a
Thus, allowing the unbanked sender to deposit money into a recipients account at any bank
branch would significantly increase network complementarity: it would reduce the access
constraint for the sender and so allow the recipient to open an account with institutions that have
rural but not urban distribution.
1 Density refers to quantity of access points and complementarity refers to the extent of the interconnection between
different bank’s networks.
An equally important challenge in the domestic money transfer business is to increase the desire
of the banks to provide transfer services or maintain accounts for low income individuals. The
costs of providing these services are directly related to the level and extent of regulation
governing account opening and the acceptance of deposits. The implementation of the
regulations governing these activities under the new FICA legislation would make the provision of
these services considerably more costly and thus less attractive to the banks. These regulations
should be reviewed and appropriate exemptions made.
There remain considerable obstacles to the provision of cost effective cross border money
transfer services for immigrants wishing to remit small amounts or to make remittances to non-
banked recipients. These obstacles present a challenge to the authorities and to the banks.
Firstly, the cross-border integration of the banking systems between Lesotho and Swaziland is an
area where regulatory reforms should dramatically reduce the costs that individuals incur to
transfer funds across the border. There seems no reasonable explanation why a transfer to
Ladybrand (SA side of the border) and Maseru should differ in cost by a factor of 7.
Secondly, money transfer product providers (for instance Western Union) do offer an important
service to poor people in that they:
§ Have good distribution in all countries that are important to SA based migrants.
§ Provide instant transfers (the recipient can collect as soon as they receive the
§ Had established distribution infrastructure in SA which was more closely aligned with the
needs of the target market than traditional banking infrastructure.
There is no obvious alternative but to encourage the development of a commercial money
transfer services in South Africa, of which Western Union is the most obvious. The regulators
need to explore ways of allowing money transfer companies to operate profitably in South Africa.
This is however made very much more difficult by the increasing burden of compliance that falls
on an authorised dealer. Furthermore if the implementation of netting agreements were the cause
of Western Unions exit, these regulations should be reviewed. This regulation probably
unnecessarily increases the cost of doing business in South Africa and should be an early
candidate for further exchange control relaxation. Alternatively, in light of the important social