August 4, 2005
I wrote a concept note for a mobile banking solution in 2005. It has many similarities to MPesa, which was launched in 2007. Although my idea never spread beyond a few emails (and I left the microfinance space in 2006), I sometimes wonder if there’s a parallel universe where I’d gone after this idea full-time.
Here’s the original whitepaper:
There is an important distinction between the methods of service delivery employed by microcredit and microsavings programs worldwide. Typically, in microcredit, branch offices manage the daily activities of credit – client selection, disbursements, repayments, etc. – and these branch offices report to and are overseen by one or more central offices. Part of the reasoning behind this structure is that clients are better served if they don’t need to travel far (or at all) from their homes and businesses to utilize credit services and that organizations are more effective when using a decentralized structure to serve their clients. Limiting transactions to a single branch office also allows for a high level of interaction between loan officers and their clients, which is considered crucial in the typical microcredit scheme. One consequence of this approach is that there are very few existing, non-bank MFIs with the capacity to disburse loans and receive repayments at more than one branch office. At the same time, though, such a service would not integrate well with several of the most common lending methodologies in the industry (e.g., solidarity groups, village banking, etc.).
This kind of ‘branch autonomy’ should not be the case with savings. There must be a clear, financial incentive to get a poor person’s savings out from under the mattress; or, there must be more than one mattress. A good savings program allows clients to make deposits and withdrawals at any of the organization’s branch offices. They can deposit savings in a rural branch and withdraw them at an urban branch (and vice versa). In some cases, they can also transfer funds between accounts – even between different banks. This distinction is critical in understanding the gap between the number of active microcredit clients and the number of active microsavings clients worldwide: a microcredit program can be viable in isolation (i.e., operating just as a single branch), whereas a microsavings program would be far less so. In other words, a viable savings program requires scale.
Now imagine a bank account that has no restrictions: a client can make a deposit, withdrawal, transfer, etc., of any amount, at any transaction point, at any time of day, and at negligible cost. Although this “ultimate” account may not be possible with current technology, and poor infrastructure limits (and will continue to limit) the spread of convenient, transaction-cost reducing technologies to many parts of the world, there is a compromise. I believe we have the potential (and have had the potential for quite some time) to make this compromise very small.
Particularly in sub-Saharan Africa, no other infrastructure network is spreading more rapidly than the mobile phone, especially in rural areas. The business of “community phones” (i.e., mobile payphones) is widespread, and it is well-known that a single phone does not imply that it is has a single user. As far as I know, in most developing countries mobile service providers sell prepaid scratch vouchers in various increments that users can purchase to “top off” their account balances (contractual subscriptions don’t make sense in these markets). In many of these countries, it’s also possible to transfer prepaid credit to other subscribers of a given company free of charge (so long as a minimum balance is maintained); this practice is sometimes used as an informal means of money transfer, as phone credit has direct convertibility to cash. The cash value of this credit, however, is not always fixed. For instance, in Tanzania, a Tsh. 5000 voucher card (approx. $5) may be purchased at wholesale for Tsh. 4800. The consumer price in urban areas and major towns (where there is access to wholesale providers) is Tsh. 5000. On the other hand, rural merchants often lack access to wholesale providers and must sometimes purchase these vouchers for Tsh. 5000 or more, thereby increasing the price of vouchers for rural subscribers to Tsh. 5200 and up. This markup is, of course, a natural reaction to market conditions caused by a poor transport infrastructure and low population densities in rural areas. But the key point here is that even though a Tsh. 5000 voucher always converts to Tsh. 5000 airtime credit, this credit does not always come at the cost of Tsh. 5000 cash (it can be both less and more).
The idea is to develop an alternate currency in the form of deposit vouchers. Vouchers are sold at wholesale in the same way as mobile phone vouchers. Ideally, the Bank producing the vouchers develops a toll-free, international phone number that can be used by clients to make transactions (or, a contractual agreement could be negotiated with domestic cellular providers to enable toll-free service for clients). Client accounts are managed much like credit card accounts: a long account number accessed by a user-defined password or pin number. (The account could be set up either by dialing the toll-free number and using an automated interface or by filling out a short form at a branch/partner office. Clients could also have multiple accounts, i.e., for home, business, etc.) Clients purchase vouchers from merchants in cash, scratch off the voucher number, and input it into their account. (Four steps are necessary: dialing the toll free number, inputting the account number, inputting the pin number and inputting the voucher number. The credit is then added to their account balance.) This savings credit can then be used in two straightforward ways. It can be used to withdraw cash at the Bank’s branch/partner offices (the normal way or even through ATMs), and it can also be transferred to other accounts (the recipient accesses his account and receives a message stating that he has a pending deposit of x amount sent from so-and-so; he then chooses to accept or decline). In this way, even at the most modest kiosks, a client would be able to transfer money from their account to the merchant’s to pay for purchases.
The next link is a simple one, but it is here that the true power of this technology lies; here is what distinguishes a technology of this kind from that of ATMs and POS devices, where technological investments are necessary, and infrastructure and liquidity constrain access; here is the breakthrough: anyone can be a bank teller. Your friend has $5 cash and you have $5 credit – send her the credit and you have the cash (so long as there is a phone nearby). In basic, conceptual terms, there’s nothing else to it. It’s an electronic “cash-barter” system with no need for a minimum [or, as I perceive it, maximum] transaction size.
Vouchers are important because every time the Bank sells a voucher, it has effectively received a cash deposit. Their elegance is that, for example, the Bank can sell a $5.00 voucher for $4.90 and charge a 2% withdrawal fee. So, even if someone were to buy a $5 voucher for $4.90, add it to their account, and then immediately try to get $5 cash from an ATM, they would need a minimum account balance of $5.10 to complete the transaction. Even in this rudimentary case (which defeats the purpose of buying a savings voucher in the first place), the Bank breaks even, though not in covering its operational expenses.
The manufacturing cost to produce vouchers is negligible (unlike smart cards, for instance) and they have no material value (which would prevent hoarding). Moreover, once a voucher has been sold, the Bank has no liability in ensuring that the voucher is actually used to make a credit deposit.
The Bank, however, does not make any money selling these subsidized vouchers nor from withdrawal fees. Although deposits would certainly be invested, the Bank makes its real money by charging a tiny transfer fee (a flat percentage of the transfer amount). Once a voucher has been converted into credit, it takes on a life of its own, going from account to account in exchange for goods and cash. There would be little need for clients to ever withdraw money directly from the Bank, for credit is as good as cash to anyone with an account.
Vouchers are not essential, however, in developed economies. These accounts could be easily linked to existing networks, including credit cards and internet banking.
This should not be labeled a pro-poor technology. The microfinance industry speaks of building inclusive financial systems for the poor, when in practice it’s more often building special financial systems for the poor. With the increasing commercialization of microfinance, practitioners worry about ‘moral drift’, that MFIs are moving up-market by focusing more on profit margins and less on the marginalized. But this debate neglects perhaps one of the greatest victories of the microfinance revolution: that Big Banks are also experiencing a ‘moral drift’ of sorts – they’re starting to focus down-market. But no matter how these two parties collide, overlap or coalesce, mobile banking transcends the need for exclusivity altogether. The operational cost of transferring one million dollars is the same as transferring one dollar, and the financial incentives for clients are identical. Although the Bank would make more money off voluminous transfers, it would have no measurable economic justification for neglecting the poor and their micro-transfers. Quite simply, it would be accessible to anyone – anywhere in the world – who is able to make a phone call.