Banking the Poor Through Everyday Stores

This paper reviews the opportunities and strategic choices facing banks considering branchless banking options. Technology, and in particular the spread of real-time mobile communications networks, permits financial service providers to delegate 'last mile' cash management and customer servicing functions to third-party retail outlets. By making basic deposit, withdrawal, and payment functions available securely through retail shops that exist in every village and neighborhood, there is an opportunity to dramatically increase the physical footprint of financial service providers in developing countries and to transform the basic economics of low-balance savings.


120
innovations / Data Democracy as the store takes in and pays back customers' cash against an equal and simultaneous electronic transfer of value between the bank accounts of the customer and the store. The bank must put in place a technology platform, which can be based on cards or mobile phones, through which it can authorize and record each transaction in real time, thereby ensuring that all client transactions are fully funded before the customer leaves the store.
A retail store's business is to buy a stock of a good that people want-let's say rice-and then resell it in smaller amounts. Through this process, the store steadily reduces its stock of rice and increases its stock of cash. When the rice inventory is depleted, the store needs to take some of its cash and exchange it for a fresh supply of rice. Now substitute "electronic money" or "bank balance" for rice, and the same mechanics apply for a banking resell business. The store starts with a balance of electronic money in its bank account, and when a customer walks into the store to make a deposit, the store transfers electronic value to the customer in exchange for cash. When its bank balance runs out, the store cannot fund any more customer deposits and it must go to the bank to replenish its electronic money stock, using the cash it has received from customers. A customer withdrawal is the same operation, but in reverse. Thus, the only difference between the store owner and the rice merchant is that the cash merchant can buy and sell electronic money for cash; that is, he or she can make both deposits and withdrawals. The store may have taken a bank loan to purchase its initial inventory of rice or electronic money, but that's an entirely separate transaction from the subsequent retailing of either commodity.
Normal stores can thus act as cash merchants, aggregating the cash needs of customers so that only the shopkeeper, rather than the entire community, has to travel to a branch to service his account. Cash merchants handle the logistics of local cash distribution, in effect bridging the distance between the bank branch and the places where poor people live and work by acting as a kind of wholesale cash aggregation point for cash merchants. The store is paid a small commission for each transaction, making this a variable-cost transactional channel for the bank and at the same time a profitable new line of business for the shopkeeper. The bank retains full control over the customer proposition and the marketing of the service, and guarantees the safety and soundness of the IT platform and risk-management policies.

FIVE REASONS WHY BANKS SHOULD MOVE BEYOND BRANCHES
So why would successful, reputable banks want to outsource a core part of the customer interface to a shopkeeper? Here are five simple reasons: 1. Decongest branches. Crowded banking halls and long queues are a common part of customers' experience in developing countries. Banks can offer a better service to their existing customers by allowing them to conduct basic transactions at a range of local shops. In this fashion, banks can offer their customers more choices and convenience; they can make their branches more appealing for high-

Ignacio Mas and Mireya Almazán
Banking the Poor through Everyday Stores er-end customers with more sophisticated needs; and they can reduce the average per-transaction cost by shifting low-value transactions, including over-the-counter bill-pay transactions for non-customers, to a cheaper variable cost channel. Today banks see transactions largely as a cost and operational burden, while most stores would like to conduct more transactions to increase foot traffic. It's a winwin for stores, banks, and their customers. Banco de Crédito del Perú has pursued this model, having signed up almost 1,500 stores that are located within 15 blocks of their own branches.
2. Develop business in new locations. Signing up cash merchants is a lowinvestment, low-risk way to test the waters in new geographic markets. It allows banks to acquire a customer base and transaction volume that, with time, may warrant the opening of a bank branch. This is analogous to the way the Opportunity International Bank in Malawi has used appropriately equipped trucks as "mobile branches" to identify favorable branch locations.
3. Create a transaction-based proposition targeting poorer segments. In order to serve poor people profitably, it is not enough to move their transactions to a lower-cost channel; it is necessary to identify an appropriate revenue pool. Traditional bank pricing models are not well suited to the poor, as they typically rely on interest margin (unattractive to the bank for small-balance savers, no matter how cheap the transactions become), account maintenance fees (unattractive to customers who do not want to commit to a fixed cost), or cross-selling credit (not relevant-or advisable-for many poor people). Poor people may not have a lot of money to save, but they have plenty of transactions to undertake: frequent small deposits building up to a savings objective, microloan installments and bills to pay, remittances among family and friends who support each other. While banks today see such transactional needs primarily as a cost-and in fact they often discourage additional transactions-these transactional needs could in fact drive poor people's willingness to pay. Per-transaction pricing can thus complement per-transaction cash merchant commissions as a sustainable banking model for poor people. The cash merchant channel could be associated with a distinct proposition for poor people. Equity Bank in Kenya has been at the forefront of transactional charging models appropriate to the needs of the poor.
4. Refocus branches on selling rather than handling cash. A branch's main role should be to sell to new customers and to propose new products to existing customers. Instead, branches are often overwhelmed by a volume of cash transactions that leave little room for a sales pitch when a customer (or non-customer) reaches the teller. By separating the service channel (delegating it to cash merchants nearer to where people live and work) from the sales channel (focused on branches with appropriately trained staff), banks may find their customer relationships significantly strengthened. Customers visiting branches could then be engaged in a conversation around their broader needs, giving them a fuller sense that the bank cares for their welfare. This is the principle behind Kshetriya Gramin Financial Services in India, which applies a private banking model to the poor and trains its branch staff as "wealth managers." In Colombia, BanColombia has developed a innovations / volume 6, number 1 122 innovations / Data Democracy parallel sales force that supports the acquisition of new customers through its 770 cash merchants.

Fill the competitive vacuum that others might otherwise fill.
Mobile operators in many developing countries have identified a clear business opportunity in offering store-of-value and payment services to their customers. Safaricom's success with m-PESA in Kenya has shown that mobile operators have the capacity to bring a compelling financial service to the mass market. Mobile operators have implemented a secure authentication mechanism with their customers (through the SIM card) and can use their own network for the secure conveyance of messages representing transactions; they already have a high-volume, transactionbased business model; they already operate prepaid account platforms and an extensive network of retailers offering prepaid (deposit) products; and they have a widely recognized brand. All they need is the regulatory leeway to operate an openended store-of-value mechanism, the appetite to incorporate mobile accounts as a core part of their business offering, and significant up-front marketing investment. Still, mobile operators that choose this course face a daunting task in building a cash merchant ecosystem with a national footprint, and most have experienced difficulty in doing so.

FIVE STEPS TO OFFERING BANKING SERVICES THROUGH CORNER STORES
A bank that recognizes the potential benefits of this banking model can embark on the approach with relatively low risk and cost. Unlike mobile operators, banks can exploit the opportunity of deploying cash merchants incrementally, since they already have an existing product range, customer base, marketing channels, and brand. A bank could start by signing up a few cash merchants around a few branches, and over time build a substantial base of merchants. For mobile operators, the proposition needs to be rolled out at scale from the start for it to have any chance of success. Banks should take advantage of the business model flexibility they enjoy and experiment with cash merchants as a way to reinvent themselves. Banks that do not target mass market financial services through retail channels may find that mobile operators have done the job for them. There are five major elements banks need to consider in their cash merchant roll-out strategy: 1. Brand protection. Trust and good service are fundamental brand values of a bank. Banks appointing retail stores as cash merchants are not only delegating the handling of customer transactions to third parties but are placing their brand logo in an environment they don't directly control. They must find ways to ensure that trust is in no way undermined. In return for the use of the logo and access to its transactional platforms, banks must insist on proper contractual terms with cash merchants. The contract should define the roles and responsibilities of both bank and store. The store must be required to maintain confidentiality of customers' transactional data, and to post adequate information so that customers are

Ignacio Mas and Mireya Almazán
Banking the Poor through Everyday Stores informed of services available, fees, customer complaint procedures, and the like. The bank will need to supervise compliance with contractual terms and monitor the levels of customer service attained, both through frequent scheduled visits and mystery shopper programs. The bank must also put information systems in place to track reported incidences at cash merchants, and must be prepared to take prompt action (including contract termination) when a pattern of misbehavior or abuse emerges. Beyond ensuring the safety of transactions, managing store branding and agent processes is essential to maintain a consistent customer experience and build trust in the new channel.
2. Incentivizing the channel of cash merchants. Banks need to incentivize the cash merchant channel adequately in order to ensure that the stores maintain a proper level of liquidity at all times, visually protect the bank's brand presence within their store, actively promote the service with the public, and take the time to train customers unfamiliar with electronic transactions. If the bank has a transaction-based revenue model, customer transaction fees can be used to compensate the cash merchant. Increased foot traffic and branding advantages from this new line of business may help the store, but it will still expect to be directly and sufficiently compensated for the service. As a rule of thumb and using round numbers, we figure that, in order for the system to work on a modest per-transaction fee of 10 cents, a typical store in a developing country might need some 50 transactions daily, producing daily revenue of $5. This would cover one person's salary of $2-$3 per day, the higher working capital requirements and risk associated with its liquidity holdings, and more frequent trips to the bank.
3. Channel management hierarchy. Banks will need an effective channel to select, sign up, train, manage, and monitor a rapidly expanding network of stores. One approach is to manage the new merchant channel in a decentralized fashion through the existing branch management structure. In this case, regional and branch managers' targets and reward structures will need to incorporate adequate performance measures for the new cash merchants. If the bank does not have enough staff at branch level to be able to take on this additional role effectively, a second approach is to have the new cash merchant channel managed by a centralized team set up in parallel to the branch management structure. In this case, there will have to be an effective mechanism for capital allocation and business planning to trade off the capabilities and aspirations of the competing channels. Alternatively, the management of the cash merchant channel itself may be outsourced to one or more third parties. In this case, the bank will need to ensure contractually sufficient control over the character and growth of the channel.
4. Choice of technology platform. The technology platform must provide for the proper authentication of transacting parties, ensure the confidentiality and integrity of transmitted data, and attain a high degree of reliability in a large-scale, distributed deployment. It must also present a convenient user experience with an intuitive customer interface. A fundamental choice is whether to issue point-ofsale (POS) devices to stores and cards to clients, or whether to rely on mobile phones as a substitute for both. A POS/card platform may be more convenient innovations / volume 6, number 1 123 innovations / Data Democracy when clients already have cards and there is an extensive card acceptance infrastructure. Because they are a dedicated device, they provide merchants a faster, more intuitive user interface. A POS-based infrastructure can leverage mobile connectivity without requiring a special deal with a mobile operator, beyond perhaps securing volume discounts. However, a mobile phone-based scheme is substantially less costly to roll out because it leverages devices already owned by both merchants and customers, and thus is likely to be much more scalable. From the client perspective, mobile phones give users immediate access to account balances and other transactions from anywhere, which cards do not. A mobile phone-based solution, however, is likely to require a deeper negotiation with the mobile operator, for example, in order to gain access to the programming capabilities embedded in the SIM card, which provides security and usability advantages. In a few countries, a telecom-led mobile money offering may already exist, in which case it may be worth considering a partnership to tap into that service and channel, rather than starting one from scratch. 5. Account opening and product sales. Beyond providing liquidity, cash merchants will likely play a significant role in facilitating opening accounts for prospective bank clients. The new channel will not result in rapid market expansion if opening an account is onerous and limited to bank branches. It is important to make account opening as easy and immediate as possible. Going a step further, thinking through the mechanism for product cross-selling and up-selling is necessary from the start, as new product placement is a key revenue driver. Separating the sales and service channels is one of the reasons the use of cash merchants is appealing, and it is possible that banks may decide cash merchants play no role in product sales but strictly process deposits and withdrawals. Nevertheless, banks should consider to what extent cash merchants could support the product sales process, if at all. Assuming banks are not expecting the store to process credit applications under a "pull" model, will they allow stores to "push" clients to the branch or to a dedicated outbound sales force? Under a push model, a cash merchant can play a variety of roles, ranging from the heavier touch of delegated screening processes to gathering the names of interested/potential clients.
The choices banks make on the above questions should be informed by customer insights on the difficulties they face today with their existing alternatives, and their aspirations. The choices will also be conditioned by the regulatory framework and by the market structure, not only in banking but in the retail and telecom sectors. There are no universal prescriptions; each bank will have to chart its own path. Banks therefore should take the time to conduct careful piloting and manage growth in a controlled way.
Banks have substantial flexibility in how they sequence the roll-out of cash merchants, both in terms of speed to scale and geographic coverage strategy. They might start with a branch decongestion strategy, whereby a few cash merchants are placed in the vicinity of their busiest branches. Such cash merchants can be promoted in a low-key way from the branch, leveraging the strong brand and market-

Ignacio Mas and Mireya Almazán
Banking the Poor through Everyday Stores ing presence of the branch within its community. Over time, banks can extend the geographic reach of cash merchants away from branches, which will require a shift to higher profile, independent marketing channels to drive customer awareness and promote their business.
As previously mentioned, this is quite different from the situation faced by mobile operators developing a mobile money scheme, who face a high critical mass hurdle for their proposition to have any visibility and relevance in the market. While mobile operators may enjoy certain advantages-control over the network, an existing distribution network of airtime resellers, a much bigger customer base-banks can deploy branchless banking services with a much lower level of business risk.
The concept of banking beyond branches may sound simple, but embracing indirect (or third-party) distribution models requires a substantial mindshift, even for the most pro-poor institutions. Banks need to consider carefully what services the poor find truly valuable, the pricing models that will allow them to realize that value, the core competencies that are most important for them (e.g., sales versus service), and, ultimately, what they want their brand to stand for in people's minds. They shouldn't lose sight of the fact that this new channel has the potential to transform their business proposition. A pent-up demand for savings and other financial services may drive it to an unexpected scale.

FIVE WAYS TO ADAPT BANKING REGULATIONS TO SAFELY EXPAND FINANCIAL ACCESS
All of this, of course, requires friendly regulators willing to allow the development of new distribution channels in banking-the possibility of using cash merchants as well as more flexible "lite" branch models. This has to be done with due consideration of the risks posed by each type of channel, but also in a way that balances the consumer protection needs of banked customers with the interests of the millions who today are financially excluded, and hence derive no protection from banking regulations. Technology, in particular the spread of real-time mobile communications networks, permits financial service providers to delegate "last-mile" cash management and customer servicing functions to third-party retail outlets. By making secure basic deposit, withdrawal, and payment functions available through the retail shops that exist in every village and neighborhood, there is an opportunity to dramatically increase the physical footprint of financial service providers and to transform the basic economics of low-balance savings and basic transactional accounts.
Below we highlight several ways typical banking regulations need to be adapted to safely allow for these new possibilities of banking beyond bank branches: 1. Retail outlets engaged as cash merchants. Retail outlets acting as cash merchants on behalf of banks need not create financial risk for either the customer or the bank, as long as the system operates on a purely prepaid basis-that is, if all customer transactions are undertaken against the store's own account and transinnovations / volume 6, number 1 125 126 innovations / Data Democracy actions are authorized in real time by the bank. In this case there is not much reason for regulators to have to prescribe which stores can and cannot be cash merchants, much less to authorize them individually.

Ease of account opening.
There must be tiered Know Your Customer (KYC) requirements, which allow for the immediate opening of small, entry-level accounts at authorized retail (non-bank) outlets. All the barriers should not be right up front for customers who are new to banking. As customers develop bigger and broader financial needs, they can be asked to perform progressively tougher KYC tests to "graduate" to more sophisticated accounts.
3. Outsourcing of bank functions. Banks should have substantial freedom to outsource components of service and infrastructures to third-party non-bank players. Scale will only be achieved through specialization, and a bank should be able to count on specialist players to conduct those functions that are not within its core competencies. Such outsourcing arrangements should preserve contractually the right of banking authorities to effectively supervise assets and activities that are relevant to ensuring the safety and soundness of the bank.

Testing and learning.
Since it is still early in the development of branchless banking service models globally, regulation should focus on enabling commercial experimentation with a range of new models while ensuring that basic, commonplace banking risks are addressed. Regulations should not prescribe business models, technology choices, or pricing structures.
5. Risk-based branching regulations. One final thought: despite the "branchless banking" label, regulations still need to be friendly to the continued penetration of branches in the territory. While retail outlets may handle the bulk of cash transactions on behalf of a bank's poorer customers, these outlets will still need somewhere to go to in order to deposit excess cash and access liquidity. In the new cash ecosystem, retail outlets handle the last mile, but banks still do the long haul. Bank branches will thus retain a role as cash distribution nerve centers in support of the bank's non-bank retail outlets located in their catchment area. Regulators should thus consider developing differentiated branch requirements that allow for "lite" branches.

LATIN AMERICA FORESHADOWS THE FUTURE
There is a growing list of pioneering banks that are successfully rolling out cash merchants, especially in Latin America. The experience in Brazil has been the most celebrated, as banks have employed alternative service delivery channels for a decade, and so-called banking correspondents have become an intrinsic feature of the Brazilian banking landscape. More than 170,000 retail outlets covering every municipality in the country bear a bank's logo, process bill payments, and often broker credit and insurance. However, the number of correspondents offering account opening, deposit, and withdrawal transactions to the poor is still low: only about 52,000 of these are authorized to perform full cash-in/cash-out transactions on behalf of banks. The majority of these full-service outlets are cash merchants

Ignacio Mas and Mireya Almazán
Banking the Poor through Everyday Stores for the three banks with mass-market strategies: Banco do Brasil, Caixa Economica Federal, and Bradesco. These cash merchants are by and large economically viable, conducting on average some 60 transactions per day, at an average commission to the retailer of 10 cents U.S. per transaction. As banks use a variety of business models, it is difficult to generalize about earnings patterns. Nevertheless, it is clear that correspondents have been a low-cost means for banks to expand their number of outlets. Caixa Economica, for example, has more than six times the number of cash merchants as branches. Moreover, new data show that the number of banking agents offering full cash-in/cash-out transactions has grown more than 40 percent in the last 18 months. If these trends continue, Brazil's network of banking agents may finally become the recognized channel for financial inclusion that it set out to be. 1 Several other Latin American countries are pursuing banking beyond branches strategies, most notably Peru, Colombia, and Mexico. With more than 5,800 cash merchants (with Banco de Crédito del Perú in the lead), Peru has the second most mature cash merchant channel in Latin America, after Brazil. However, it also has its limitations. Many banks in Peru are primarily setting up agent arrangements as a branch decongestion strategy, rather than to reach new market segments. This is because account opening at cash merchants has only recently been allowed by regulations The experience in Colombia is less developed, but it already gives us an indication of how transformative these new business models can be. Colombia's banking regulations allowed for cash merchants beginning in 2007, with Mexico following two years later. BanColombia, the largest bank in Colombia in terms of assets, deposits, and number of customers, has approximately 770 cash merchants performing a healthy number of transactions (40-60 transactions per day). A parallel, roving sales channel-Puntos de Atención Móviles (PAMs)-has been deployed by the bank since 2003, preceding the advent of agent banking regulations. Operated by BanColombia employees that set up temporary sales stations, PAMs can open deposit accounts, facilitate opening other products, and generally sell the full range of BanColombia products to potential customers. Importantly, PAMs do not process transactions; cash-in/cash-out is done either at the branch or at a cash merchant. As of December 2010, BanColombia has over 550 PAM points served by around 200 BanColombia employees, which complement and support their cash merchant roll-out strategy. 3 In Mexico, the experience is still nascent. In late 2009, Banco Walmart became the first bank in Mexico authorized to operate through retail agents (in their case, 1,400 Walmart stores). Roughly a dozen banks are now authorized to operate through agents, including Mexico's largest bank, Bancomer, with the national telegraph service. The national savings and development bank, BANSEFI, is currently partnered with a network of rural community-owned stores (Diconsa) to offer basic transactional accounts. It is too early to tell what these partnerships will yield in terms of financial deepening, but it is encouraging that numerous retail banks are capitalizing on the opportunity. 4 innovations / volume 6, number 1 127