I had the great good fortune on Tuesday to sit in on the first day of the annual conference of the Institute for Money, Technology and Financial Inclusion. IMTFI is housed at U.C. Irvine and led by anthropologist Bill Maurer. The mission is to support research on money and technology among the world’s poorest people: those who live on less than $1 per day. The Institute is funded by the Bill and Melinda Gates Foundation through their Financial Services for the Poor program.

IMTFI is building a network of researchers, often from smaller universities in the developing countries, who are doing a wide range of projects. It won’t surprise anyone to hear that a lot of the research revolves around the use of mobile phones for financial services. Mobile wallets are, in a variety of different configurations, providing access to bank account-like features (savings, transfers) to previously unbanked people. These wallets are sometimes being offered directly by carriers, by carrier-bank JVs, by other non-banks and (in a few cases) by banks themselves. The poster child of this world is the M-PESA product offered by carrier Safaricom in Kenya.

Mobile wallets are springing up all over – literally it seems as if a mobile wallet scheme is either in the market or in some stage of planning in every country. It was fascinating to talk to researchers who were looking into a variety of cultural, economic, regulatory, and technical challenges with these wallets.

Being exposed to the sheer volume of activity in mobile wallets in the developing world made me think more about cross-border remittances and the role these wallets will play in this sector of the payments industry. I think the answer, over the long term, has a lot to do with the growing use of a “decoupled” payments model.

Decoupled Defined

At Glenbrook, we’ve been talking and writing for some time about the “decoupled” model in the payments industry. We’ve used the term to mean a situation where a payment provider does two payments transactions (one between the payer and itself, another between the payee and itself) instead of facilitating a single, “end to end” transaction.

Consider, as an example, a check written from a consumer to a biller – that’s an “end to end” (not decoupled) transaction. If the same consumer goes to their online bank service, the bank does one transaction (a book transfer) to take money out of the consumer’s account and credit its own with itself. The bank then does a second transaction (either an ACH or a check) with the biller: the consumer is not a party to this second transaction.

Although it seems like this two-transaction model would be more complicated or costly, in fact it turns out that the two-transaction model offers the provider (the bank in this case) a lot of flexibility. It can use different systems for the two transactions; it can vary timing to manage risk; it can aggregate one side of transaction but not the other. It’s a great model. PayPal uses this model; so does Western Union; so does Fiserv’s ZashPay (and most P2P schemes in the market).

Cross-Border Decoupled

So why am I talking about decoupled payments transactions in connection with cross-border remittances? Because I think a version of this concept will work very well applied to these payments. Cross-border remittances are mostly migrant workers sending money home. The flows are talked about in terms of “corridors” – as any sender of money is normally only sending money to one person, in one country. The send-country/receive-country pair is the corridor. Historically, cross-border payments services targeting these corridors needed to solve the whole problem – collecting money from the sender and getting money to the receiver. They did this, again historically, through a complicated set of agent relationships and/or deals with local banks or networks.

These transactions were generally executed on a decoupled basis (you can argue, in fact, that all cross-border transactions are handled on a decoupled basis). But what I anticipate is that going forward, the actual business relationships will increasingly be decoupled as well.
Imagine you are starting a payments service in a developed country to help migrant workers send money home to India. You need to invest in market outreach to find the workers and figure out the ways you are going to enable them to fund their transactions. But rather than needing to invest in an agent network in India, you can simply negotiate a deal with the Indian mobile wallet provider(s) that are serving the Indian market, and deliver the transactions into that wallet. It will take time, of course, to work out the business structure of these deals (who pays whom? who does the FX?) but operationally, it should work really, really, easily. Furthermore (and I’m going out on a limb here) I’d say that over time, regulators should embrace this approach, as it should reduce risk and cost to the consumers on both sides. Of course, there will need to be some know-your-customer checking (on both sides) to ensure that schemes aren’t being hijacked for money laundering or other nefarious purposes. But the good thing about the cross-border remittance market is that it is repetitive, so once the first transaction has cleared, subsequent ones should be easier.

The “decoupled business relationship” means that the receiving customer doesn’t have to know, or be involved with, the service the sender is using – and vice versa. It should just work! Long term, it may even alter the need for scale in this business – for example, a small, non-profit community group, serving migrant workers in a developed country, could, in theory, play in this game – or (perhaps more realistically) be a front-end for a larger bank or payments services provider, who negotiates the deals with the receive-side wallets.

It is interesting to reflect that it may be the activity in the developing world that spurs the developed world payments model to evolve. What do you think? Share your comments below!