This is an excerpt from a recent CGAP paper, Banking on Mobiles: Why, How, for Whom? In it, Kabir Kumar and Ignacio Mas examine the business case and deployment options around mobile banking for smaller banks and microfinance institutions. With effective partnerships and technical choices (which affect customer uptake), we believe there is a strong market opportunity to reach poor people with a broad range of financial services.

Mobile phones are ingenious devices, but one thing they cannot do by themselves is convert cash into electronic value or dispense cash. They can be used only to transfer or transform value electronically. A mobile banking platform therefore needs to be supported with a cash conversion platform—whether full-blown bank branches, ATM terminals, or third-party banking agents. Remember, the whole mobile proposition is based on choice and control: if I don’t have a choice of cashing in or out of my electronic wallet, I am not likely to think mobile banking is doing much for me.

A bank that wants to cover a new geography with a mobile banking strategy will need a cash in/cash out network in that same geography. The mobile platform will not “liberate it” from having to figure out a physical cash delivery channel: it’s clicks and mortar. But it can be “clicks and mortar lite,” because the mobile banking platform itself can help in deploying that network cost-effectively by letting, for instance, an agents’ mobile phone act as a POS for handling agent transactions .

To what extent will mobile payments replace the need for cash? To answer that question we need to understand the relationship between electronic money that may be accessible through mobile phones and cash. There are two key roles of money: as store of value and as means of payment or exchange. In a totally cash-based economy, cash fulfills the two roles at the same time. It may be inconvenient for people to store and move so much cash around, but at least there is no risk of lack of liquidity: if you have enough value stored in cash, then you have enough cash to make a payment.

It is only when the two roles are separated that liquidity may become a problem. For instance, as people move into savings in real goods (bricks, chicken, real estate) in search of convenience and higher returns in the store-of-value function, the total amount of savings far exceeds the amount of cash easily available. Liquidity might be a problem if you are not able to convert your physical savings into cash readily enough. In that case, you might need to resort to bartering your physical assets—so you tend to choose physical stores of value that are tradable and movable (e.g., livestock). When people need to resort to barter, cash has lost both functions—as store of value and as means of exchange.

Similarly, once there is electronic money, the demand for cash depends strictly on the extent to which this new electronic form of store of value can also be a means of exchange. Can we use electronic money to directly pay for all that we need? In the language of payments, is there a wide acceptance network?

To the extent that electronic money is convenient for us to use and widely accepted, we return to a world where the same “asset” is performing both functions. Now, there is no liquidity issue with electronic money and cash loses its role. But if electronic money is not widely accepted, and we know of only a handful of places where it is accepted as a form of payment, storing my value in electronic form exposes me to liquidity risk: I cannot ensure that I can convert electronic value into cash quickly enough if and when I need to spend the value.

People have heralded the end of cash before but the claims of a cashless society are more emboldened today by the growth in cell phone use globally. In fact, the relationship between electronic money accessible via mobile phones and cash is likely to vary over time.

In the early stages of introduction of this electronic money, it will be essential to ensure that people have ready cash-in/cash-out options to support the mobile proposition. Customers will want to test the liquidity of electronic money. Typically they will want to convert all electronic money inflows into cash immediately and in full. In places where cards have been used with microfinance clients or low-income customers, people who live in all-cash economies, most cash out their loans, government benefits, and so forth, on the spot immediately after they receive them.

As electronic money gains acceptance as a store of value, people might restrict getting their cash out to meet their daily liquidity needs for very small value transactions only. In fact, in this stage, we expect there to be substantial net cash into the mobile banking system as it absorbs cash that had previously been performing the store-of-value function (under the mattress). Over time, because goods, services, and salaries will be paid increasingly through electronic transfers, cash conversion will become unnecessary. Direct electronic transfer will skip the electronic value-to-cash conversion on the paying side and the corresponding cash-to-electronic value on the receiving side. Now, finally, total cash substitution is happening, both as store of value and as means of exchange. Ultimately, cash-in/cash-out points may not be necessary at all. But you would do well not to plan for that world as yet…