For some time I have been intrigued by banks' lack of interest in, or inability to leverage their ATM networks to gain a competitive edge. Traditionally used for cash advances and balance enquiries, ATMs have turned into more of a multimedia kiosk capable of providing a wide range of services. One area of particular interest, in my mind, is money transfer, a high-margin banking service largely dominated by two non-banks, notably Western Union and MoneyGram.

I was therefore pleasantly surprised to discover a solution at the recent Cartes Afrique conference in Tunis that exploits a bank's ATM network for such purposes. A Tunisian bank proposes a simple, innovative, cost-effective service for local money transfer. In simple terms, it works as follows:

• The sender initiates the transaction in the bank branch, in the online bank, or at the ATM. In the branch, the sender may pay by cash or card, in the two other channels the sender uses a card (any bank card) to pay for the transfer.
• At the point of interaction, the sender selects a unique PIN code for the transaction, and registers the recipient's mobile phone number as part of the money transfer transaction.
• The sender then contacts the recipient (typically by phone) to communicate that a money transfer has been initiated, and the PIN code for the transaction.
• To receive the money, the receiver simply goes to one of the bank's ATM, and keys in his mobile phone number and the PIN code for the transaction. The information is then verified by the bank system, and if correct, initiates a call to the mobile phone. The customer is asked to accept the call and place the phone in proximity of the ATM's loud speaker. The ATM then sends an encrypted signal to the mobile phone to authenticate the mobile phone.
• Upon verification, the cash is dispensed to the receiver as a regular cash advance.

The advantages of this solution are several:

• The bank may leverage its existing infrastructure to propose a new service, and generate additional income.
• The service may be proposed to non-bank customers, as even customers without a card or bank account may utilize the service. All you need is cash and a mobile phone. It can even be leveraged as an acquisition channel for new customers, and you already know their phone number, so it is easy to make contact.
• The service is secured using two factor authentication at point of reception. The receiver can only access the money if (s)he is in possession of something(s)he knows (the PIN), and something (s)he owns (the mobile phone).

One area to look into in more detail is how the model stacks up against the relevant market's anti-money laundering (AML) regulations. However, I personally don't see too many barriers in this regard. The sender is known, as the transfer will have been paid using a card (online or in ATM), and thus traceable, or over the counter in the bank branch, with the possibility to ask the sender to provide identification. On the receiver side, the receiver's phone number is known, and the user may thus be known. However, there are of course ways to obtain an anonymous, prepaid mobile phone number. Some additional measures should however mitigate the risk of large scale money laundering.

Such measures would include setting transaction limits on each money transfer transaction, monitoring velocity, and track abnormal transaction activity on both the sending and receiving side of the transaction. Excessive use of the service by a sender or a receiver may subsequently be suspended or blocked. Limiting the service to small amounts –the average money transfer transaction from France to African destinations is in the €200-300 range, as an example – combined with the need to be physically present at the ATM with a mobile phone to receive the funds, would make the service unpractical for money laundering of large sums of money.

But the service would still be highly attractive for customers in the target segments wishing to transfer funds to friends and family.