AFI’s first annual Forum took place in Nairobi, 14-16 September 2009, and was co-hosted by the Central Bank of Kenya. The Forum had two objectives:

1) To build a community of policymakers who can share their collective knowledge on policy solutions to promote financial access for the poor.

2) To draw on the collective knowledge of members and strategic partners, including researchers, donors and private sector partners to create a practical roadmap to create policies which promote financial inclusion.

From the perspective of mobile money for the Unbanked, the key take-aways from the event were:

• Financial regulators in developing countries now have a forum which allows them to discuss ways to accelerate financial inclusion in a coordinated manner. This forum also enables regulators in developing countries to learn from their own ‘champions’ – regulators who are most successful in accelerating financial inclusion.

• When looking at the most successful countries for mobile money, the ‘test & learn’ approach to regulation is emerging as the single most important condition to creating an enabling regulatory environment.

• Competition is an important dynamic to foster innovation, choice and cheaper prices for consumers – all of which are highly beneficial for achieving financial inclusion. There is still some way to go to convince financial regulators that there are more ways to promote financial inclusion than the traditional ‘bank-led’ business model.

• A lot of research is still necessary to inform policymaking for financial inclusion. However, more emphasis should be given to the needs of consumers.

A detailed description of the event is provided below:

***Day 1***

The first day of the Forum focused on key financial inclusion policy challenges. Opportunities and barriers to progress were discussed, including the regulatory hurdles and additional data and survey methodologies needed to make the right strategic choices. Professor Njuguna Ndung’u, Governor of the Central Bank of Kenya opened the forum and was followed by various speakers who provided regional overviews of the state of financial inclusion in Latin America, Africa and Asia.

This was followed by a discussion between researchers and policy makers on survey methodologies and on the benefit of using survey data for financial inclusion policy design. The discussion revealed that two approaches are currently being used to designing financial inclusion policy. The first approach is the ‘top down approach’, which involves creating a high level approach and applying it to the market. The second approach is the ‘bottom up approach’, wherein the needs and behaviour of the consumer are understood as a first step to informing policymaking.

From a GSMA perspective we would very much hope that policymakers use both approaches, bottom-up (to consider the actual needs of consumers) and top-down, at the same time. It is still too rare that regulators apply the ‘test and learn’ approach, where policy solutions are created by following the developments of the market and the actual consumer demand. The relatively low success of providing financial inclusion so far may be accounted for by policies which have not really taken into account the needs of consumers.

***Day 2***

Day two was a very interactive day. Roundtable discussions called ‘fishbowls’ focused on AFI’s six policy areas: mobile banking, financial identity, agent banking, state bank reform, consumer protection and diversification of financial products and providers.
Each ‘fishbowl’ consisted of 6-8 panellists who tackled the respective issue in a moderated discussion. After an initial discussion between the panellists the discussion is then opened to the audience.

Mobile Banking ‘Fishbowl’ Discussion

The Panel was moderated by Michael Tarazi, Senior Regulatory Specialist, Consultative Group to Assist the Poor (CGAP) and the panellists were:
• Nestor Espenilla, Deputy Governor, Bangko Sentral ng Pilipinas
• Gerald Nyaoma, Director of Banking, National Payment Systems, External Payments and Reserve Management Department, Central Bank of Kenya
• Carlo Corazza, Payment Systems Development Group, Financial and Private Sector Development, World Bank/IFC, USA
• Tavneet Suri, researcher at MIT, USA
• Marina Solin, Programme Director, Mobile Money for the Unbanked, GSMA, UK
• Abdul Qadeer Fitrat, Governor Da Afghanistan Bank

Nestor Espenilla and Gerald Nyaoma started the discussion by sharing their experience with enabling mobile money in their country. Mobile money in the Philippines and Kenya has been so far most successful in terms of business development and consumer adoption. It became clear that both regulators achieved success by using the ‘test and learn’ approach. This means that there was an initial period of dialogue and mutual learning between the regulators and mobile operators. The mobile operators were allowed to experiment and to develop their business model under close supervision of the regulators. This created a fertile regulatory environment, where market players are closely supervised without initial regulation specific for mobile money. Once the learning and innovation in the market had been implemented and satisfied the needs of regulators and mobile operators, both regulators envisaged or implemented regulation to provide legal certainty to the status quo and to open the level playing field to new entrants. The Philippines have recently adopted an – e-money Circular which regulates non-banks when offering e-money. The Kenyan policymakers are currently working on a new payments law, which will be open to consultation once ready.

This segment was followed by comments from:

Carlo Corrazza, who made a statement to promote interoperability for mobile money. This should initially not be mandated, but mobile operators should be ‘called to the table’ to make a commitment early on in the process.

Tavneet Suri, who explored the possibilities of research to inform policymakers about customer needs. Much still needs to be learned about consumer behaviour, characteristics and needs.

Marina Solin, who explained the advantages mobile operators bring to financial inclusion. Mobile operators have a much better reach to unbanked customers than financial institutions, due to the reach of the mobile distribution channels, mobile penetration and better brand recognition by unbanked customers. The competitive nature of the mobile industry is also a good influence for promoting more services at cheaper prices to consumers. Whilst ‘there is no right business model’ to promote, it is important to allow market players such as banks and mobile operators to negotiate cooperation on commercial terms. Several business models are likely to emerge and they should compete with each other benefiting the consumer by offering a better choice of cheaper services. If one player, for example the bank, have a regulatory advantage for the whole range of financial services offered to unbanked customers, then the business relationship between the bank and the mobile operator is likely to be skewed. This can limit innovation and prohibit choice and cheaper prices to consumers, because the bank has the incentive to extract too much value from the cooperation and also to impose old ways of doing things. Commercial negotiations and healthy competition should lead to innovation under close supervision of financial regulators until it becomes clear which services benefit the consumers most. Formal regulation should then follow the market.

Abdul Quadeer Fitrat, who talked about mobile payment services and the regulatory challenges in Afghanistan. In his view pushing out the ‘frontier of access’ with mobile banking can increase poor people’s access to credit, increase family savings and decrease loss of money. Whilst these benefits need to be harnessed the Central Bank has to ensure that risks such as identity theft, money laundering and financing of terrorism need to be mitigated.

***Day 3***

Day 3 consisted of an in depth demonstration of the M-Pesa service in Kenya. The audience learned how the service developed, what benefits it brings to consumers, how to register and use the M-Pesa service. M-Pesa is now the most successful mobile money service in the world with 7 million customers and 11,000 agents.

Atiur Rahman, governor of the Bangladesh Bank spoke about financial inclusion as a tool for combating poverty in Bangladesh. It was interesting that the governor initially encouraged central banks to encourage ‘bank-led’ mobile banking solutions, because this provides more certainty and clarity for central banks that they can regulate the mobile banking solution. However, in the discussion, he conceded that there may be an opportunity to experiment more with mobile operators in the future.

Please feel free to share your comments or thoughts below.

On 1 July 2009 a new amendment to the Regulation of Interception of Communications and Provision of Communication-related Information Act, 2002 (RICA) came into effect. RICA is intended to assist law enforcement agencies with the tracing of criminals where mobile phones are used to commit major crimes.

The new RICA provisions prohibit mobile network operators from activating a new SIM card unless they have captured the customer’s identity. This includes the person’s cellular phone number, full names and surname, identity number and an address (preferably a residential address). The personal information of the customer must also be verified by means of a current (green bar-coded) national identity document, a temporary identity certificate, a valid passport or a travel document. The operator is also required to retain a certified photocopy of the relevant identity document.

Mobile network operators such as CellC, MTN and Vodacom have to comply with these provisions since 1 July 2009 whenever a new mobile phone number is activated. However, the measures also extend to all existing customers. These customers have 18 months from 1st July 2009 to register both their prepaid and contract SIM cards in accordance with the new provisions. SIM cards of subscribers that fail to comply with RICA within the specified time period will be disconnected from their networks until they are registered.

General Consequences
Customers in South Africa could buy a prepaid mobile phone without the need to register. 86% of mobile subscribers in South Africa have such a prepaid phone. RICA’s new requirements bring the considerable task of registering the existing 86% of prepaid customers and the requirement to capture customer identity and proof of address. This has already resulted in a drastic reduction in the number of new customer registrations for mobile services in South Africa.

In addition, concerns have been raised in the press regarding the efficacy of the new provisions in fighting crime, the levels of cooperation that could be expected from customers, and the possible abuse of personal data in identity fraud.

Consequences for mobile money for the Unbanked
South Africa has been creative and innovative in balancing financial inclusion and crime combating. It softened the negative impact of its money laundering laws on financial inclusion by creating proportionate exemptions for low-value mobile money services.

This means that a customer in South Africa can register for a mobile banking service by opening their bank account with their mobile phone. According to Guidance Note 6 (known as ‘Circular 6′ on mobile banking) and FICA Exemption 17, there is no need to go to a bank branch initially if the customer has a valid South African identity number and if transaction limits are observed. The customer can therefore start using the mobile banking service by transacting small amounts without going to a bank branch to provide an address. This approach is proportionate to risk, because the identification requirements become more onerous as the transaction sizes increase. If the customer wants to transact higher amounts, a full identification and proof of address has to be provided in person either at the mobile operator or abank representative (depending on the amounts transacted).

The new RICA rules are undermining some of the space that these exemptions created for mobile money services. RICA re-introduced some of the classic financial inclusion barriers by linking the heavy identification requirements of RICA to the right to use a mobile phone. Whilst customers have reduced identification requirements for low-risk mobile banking services, they now have to undergo the more rigorous RICA procedures to be able to use a mobile phone. This negates the benefits of Guidance Note 6 and Exemption 17.

Clients will, for example, face geographical barriers because they will need to visit agents of operators in person where they will be required to produce the required documents. Many will probably find it inconvenient or expensive to meet the requirements while other users may lack the required documentation. Those who lack identity documents are often socially and financially vulnerable. A significant number of South Africans will therefore be unable to access mobile communication services and therefore also mobile money services.

All the above difficulties will exclude many South Africans from mobile communication services, not because they are criminals, but because they have a hard time obtaining the relevant official identity document. This will impact their ability to use mobile phones as a communication tool. Many low-income South Africans conduct their trade and business via their mobile phones. In addition to the social impact of losing their mobile connection, it will also impact on their ability to continue to earn their income as builders, painters, plumbers, etc .

The exclusion will potentially impact third-party service providers as well. The RICA provisions can potentially close down one of their primary means of staying in contact with a client via the mobile phone. In particular this could lead to greater defaults (in loans, utility bills, etc) as such a client is no longer reminded of, or chased for, outstanding payments.

Conclusion
It is too early to predict with certainty the outcome of the new RICA requirement on mobile money in South Africa. However, it is likely that the new barriers will prove too high for some current and potential users. It is, however, possible that users will go to great lengths to meet the new requirements because mobile phone connectivity is so important in their lives.

Once the hurdle RICA presents has been surmounted, there may even be a positive impact for mobile money services. If the identification processes of RICA are robust enough and provide sufficient proof of the identity of the user of a SIM card, it may justify lesser money laundering controls in respect of mobile money services rendered via that SIM card. Those who meet the new requirements may then benefit from a further relaxation of the money laundering laws relating to mobile phone banking.

South Africa therefore provides an interesting case study that merits close attention as it unfolds over the next few years. 

Notes:

South African Reserve Bank, (2008), Guidance Note 6/2008 Issued in terms of Section 6(5) of the Banks Act, 1990: Cell-phone Banking, South African Reserve Bank, Pretoria (previously issued as Bank Circular 6 of 2006.) and FICA (South Africa) FIC Exemption 17 and guidance notes; available via www.fic.gov.za

Louis de Koker’s letter to the Select Committee on Security and Constitutional Affairs on 31 October 2007.

In my opening remarks, I emphasized that “we don’t have to agree, but we can all listen to one another”. What followed was a great deal of productive dialogue. The group took advantage of this unique opportunity to pursue their curiosity about the topic of mobile money and engage in some frank and open discussions. There were are a few key themes and statements that I would like to highlight from the discussion..

Unpacking mantras. The group had a strong desire to explore just what phrases like ‘proportionality’ and ‘financial inclusion’ really mean. I saw a lot of value in discussing the practical implications of these mantras and the role that they actually play in regulating mobile money. Further, hosting this discussion at a global level enabled participants from one market to evaluate the attitude and approach used in other markets.

Recognizing the value of dialogue. Operators shared experiences engaging with their regulators and noted that when dialogue takes place, learning goes both ways. As one participant put it: ‘several years ago, our regulator taught us about the need to identify customers. This requirement is now common in telecom and we were effectively prepared for this change when it arrived based on our early experience in mobile money. Our regulator also required that we monitor all transactions and report on what is happening to satisfy AML requirements. To satisfy this requirement, we ended up creating a very powerful management information system that now enables us to make better business decisions. Likewise, we helped our regulator understand some commercial implications of regulation. For example, the relationship between stringency of KYC and ability to scale, and the impact of account balance limits on targeting prospective high value customers.’

Above all, both operators and regulators emphasized a desire to keep dialogue open.

‘Identify the risk and regulate it proportionately’. This phrase was initially used during the Mobile Money Summit, but panellists and participants returned to it during the Leadership Forum. The phrase is simple, but it effectively communicates the approach used in Kenya – both prior to and following the launch of M-PESA – and could be considered as an effective philosophy for regulating mobile money.

I would like to sincerely thank the regulators and mobile operators who participated. I look forward to seeing you again next year.

What we like about South Africa’s AML/KYC regulation

Mobile experience and proportionality
A customer in South Africa can register for a mobile banking service with a truly ‘mobile experience’ by opening their bank account with their mobile phone. There is no need to go to a bank branch initially if the customer has a valid South African identity number and if the following limits are observed:

• daily transfer limit of approx. US$100 (approx)
• monthly transfer limit of approx US$2,500 (approx)
• maximum balance of US$2,500

The customer can start using the service by transacting small amounts without going to a bank branch
to provide an address.

This approach is proportionate to risk, because the identification requirements become more onerous as the transaction sizes increase. The customer has to provide identification when transacting up to US$500 with the same monthly limit and maximum balance of US$2,500 respectively. If the customer wants to transact higher amounts, a full identification and proof of address has to be provided in person to a bank representative.

The appropriateness of the actual daily/monthly transfer limits as well as balance limits may depend on the risks of the service and on the customer group. In addition, different transaction limits may be appropriate in different markets. However, the underlying principle of low transaction sizes constitute low risk and should be less onerously regulated than higher transaction sizes which constitute higher risk, is what is key for a proportionate regulatory solution.

Potential for improvement
This solution is limited to customers who have a South African identity number. This means that migrant
workers do not qualify.

Question for Discussion: which elements of South Africa’s AML/KYC solution can other countries replicate? Which regulatory solution should we focus on in our next Quarterly Update?

Openness: the extent to which new models that had the potential to be transformative were not prohibited from starting up.

Certainty: the extent to which policy makers and regulators provide clarity that reduces the level of risk for private sector operators not only at start-up but over time.

Quadrant 1 (high certainty and high openness) is clearly the most suitable situation to facilitate mobile money, hence the direction of the arrows. However, based on early diagnostic work in two African countries (South Africa and Kenya), we hypothesised that middle income countries like South Africa were more likely than low income countries to be in quadrant 2 (high certainty, low openness) because they typically have more developed regulatory regimes. That is, they usually have a range of regulatory institutions that have issued regulations or guidance on mobile money or related issues, increasing the certainty, but they face the real risk that a plethora of overlapping and sometimes obsolete regulations will reduce the space in which they can innovate. Conversely, we hypothesised that in low income countries like Kenya, there was simply less on the books in the way of legislation and regulation, usually resulting in more discretion for regulators (and correspondingly, less certainty for providers), but this lack of regulation could also create more openness for the development of innovative models. The report suggested which regulatory domains were affected by mobile money but did little to prioritise among them.

Recently, three colleagues Lyman, Pickens & Porteous (2008) went further, suggesting how to prioritise the factors based on identifying two necessary conditions for branchless banking to emerge from the range of country diagnostic missions undertaken in 2007. These conditions are:

1. Agents must be allowed to operate on behalf of banks and others to open accounts and handle cash in and out functions.As Lyman, Ivatury, and Staschen (2006) argued convincingly, this particular arrangement greatly extends the potential reach of the financial system since existing businesses, such as local merchants, can function as financial service points at much lower cost than if a bank had to set up a new branch or even an ATM infrastructure.

2. Regimes to oppose money laundering and the financing of terrorism (AMLCFT) should be proportionate. Specifically, the due diligence procedures required under Know Your Customer (KYC) regulations (now present in most countries) for opening new deposit accounts or taking payments, must allow for reduced identification and verification procedures for low risk customers. Otherwise, low income people could never meet the standards set in developed countries which, for example, require them to verify a physical address by presenting a utility or other bill.

Lyman et al. (2006) identified four more regulatory areas (“next generation issues”) that would affect the trajectory of development. There should be an appropriate space to issue e-money and other stored value instruments, along with effective consumer protection, an inclusive system to regulate payments, and appropriate competition rules for new payment systems. They also affirmed the earlier observation that, because the regulation of mobile money cuts across many regulatory domains, the risk of coordination failure is higher.

From Information to Rating

To proceed from the general insights about regulatory factors above to a country rating model for mobile money, we took two steps:

1. We designed a simplified questionnaire that collects answers about the status of policy or legislation (including regulation or guidance) across the two dimensions of openness and certainty and in the main domains bearing on mobile money.

2. We then developed a scoring model that weights the answers obtained relative to the purpose of the rating expressed above.

Initial Results

We focused on using publicly available information from diagnostics in four key countries, chosen because of their scale and the levels of interest and activity in mobile banking. They include three leading countries that are mobile money pioneers in the developing world – Kenya, the Philippines and South Africa – and India. In India, interest in mobile money has exploded in the past three years after mobile subscriptions took off in that massive country with its great infrastructural challenges for traditional models of extending the financial system. Several Indian banks have introduced various mobile channels for their customers, although mass usage of those channels reportedly remains low and is certainly not yet transformational.

While this is a small sample, these four countries offer sufficient variation in terms of income level per capita and extent of activity and development in mobile money to test our hypotheses about enabling environments and the growth of mobile money.

The outcome of the scoring summarised in Figure 2 supports several of the hypotheses I advanced earlier. Three points are especially important. First, the openness of the environment does indeed matter: in all of the three countries that are ranked much higher on the openness axis (i.e. to the right hand side), mobile money models have been relatively more active for longer and are more widely used, compared to India, which lies lower on the openness axis.

Second, however, the countries classified as middle income by the World Bank (South Africa, the Philippines and India) all lie higher on the certainty scale than low income Kenya which is clearly in the bottom right quadrant (high openness but low certainty). The Philippines is positioned just inside the top right quadrant, reflecting the fact that while its environment is very open, some of the models have been authorised based on bilateral letters of agreement, the Filipino Central Bank (BSP) is now moving beyond this level of discretion towards a broader framework in key areas like e-money issuance by non-banks. On 9th March 2009, the BSP published a guidance circular setting out an approach for nonbank e-money which increases the certainty around this issue: in fact, it results in an increase of 0.5 in the certainty score for the Philippines, decisively boosting its position within the top right hand quadrant.

In some ways, South Africa’s position is surprising: experiments began there comparatively early and several well known pioneering models such as WIZZIT and MTN Mobile Money emanate from there, but the role that non-banks can play in issuing e-money is circumscribed by the current guidance note on e-money which has frustrated some potential innovators. To further enhance its environment, South Africa would have to amend its position, for example by creating a category of regulation for non-bank e-money issuers, or ‘narrow banks,’ a step that has in fact been suggested. In general, common law countries have an advantage in terms of openness because of the presumption that whatever is not prohibited is in fact allowed; in civil law countries, the reverse applies.

Finally, India has a plethora of legislation. Its laws, regulations and guidelines across a range of areas provide certainty, but limit the openness on key issues: for example, what types of entities can serve as agents? The Reserve Bank of India first allowed agents to function in 2006, and in 2008 issued further guidelines to clarify the restrictions on this role.

While it is easy to make the case for openness, recent events in Kenya also show the importance of certainty. For example, M-PESA, perhaps the largest single mobile money model in these four countries based on number of users, is not formally regulated but operates at a system-wide scale under a no-objection letter from Kenya’s Central Bank.

Clearly, openness and certainty alone are not sufficient to ensure the sustainable development of mobile money. The safety of clients’ deposits matters too, as Lyman Pickens and Porteous (2008) point out. In the absence of a framework that creates certainty about which types of entities can enter and how they must behave, too much openness to innovative models from new entrants can be risky, especially once these models move beyond the small-scale pilot stages.

Question for Discussion: what is the right balance between openness and certainty and how can regulators find it?