The Replication Limits of M-Pesa in Latin America

20 July 2016
7 comments

Lea en español / Read in Spanish on the Microfinance Gateway.


M-Pesa ceased operations in South Africa on June 30, shedding light on an ongoing discussion: Is the challenge of replicating the success of the M-Pesa model in Kenya more about implementation and management or about context and market structure? The fact that the same parent company has not been able to replicate its success story in South Africa seems to support the latter explanation.

In Latin America and the Caribbean (LAC), we have been dealing with this dilemma for a long time, and the evidence points to context as well. M-Pesa brought innovations to East Africa that other players had already brought to most LAC markets.

Photo Credit: Marco Simola, 2011 CGAP Photo Contest

Proximity and Simplicity

Stripping it of several possible adjectives (digital, branchless, mobile, etc.), M-Pesa’s two core innovations have been proximity and simplicity. Customers have massively embraced the ability to access financial services close to home without bureaucratic barriers. This has had a fantastic impact on financial inclusion.

Large agent networks and simplified account opening processes made proximity and simplicity possible (whether regulation enabled or trailed these innovations is a different discussion). At the most basic level, low-cost communications platforms enabled this, by allowing disperse nodes in a network to transact instantly.

Nonetheless, these innovations, either in tandem or on their own, have radically changed only two types of financial services so far: bill payments (including airtime top-ups) and person-to-person (P2P) domestic remittances. According to the GSMA, these transaction types accounted globally for 96% of the volume and 87% of the value in 2015 in the case of M-Pesa-like providers.

This diversity of players delivering proximity and simplicity is a powerful explanation for the replication limits of M-Pesa across markets and regions. While successful in a handful of markets in LAC such as Paraguay, Honduras and El Salvador, the M-Pesa model could not thrive in markets where its value proposal was not better than what was already offered, or in other words, where proximity and simplicity had already been achieved in large scale by other business models (see this CGAP paper on market types for more information).

The Provider Landscape in LAC

A look at the provider landscape in LAC for bill payments and P2P gives a sense of the scale at which LAC already embraced these innovations, and the reach they have in markets with low financial inclusion.

The following table shows five countries – where two-thirds of all people in LAC live – and maps the large-scale providers of bill payments (excluding airtime top-ups) and P2P domestic transfers in each market (large-scale is defined as encompassing more than 1,000 access points and more than 1 million transactions per month in these two categories). It displays the diversity of business models in the region: mobile network operators, banks, postal services, retail chains and aggregators working with small merchants all became providers of these innovative financial services in different settings.

Country

Provider

Type

Argentina

Cobro Express

Aggregator

Argentina

Pago Facil

Aggregator

Argentina

Rapipago

Aggregator

Brazil

Banco do Brasil

Bank

Brazil

Banco do Estado RS

Bank

Brazil

Banco Santander

Bank

Brazil

Bradesco

Bank

Brazil

Caixa Economica Federal

Bank

Brazil

HSBC

Bank

Brazil

Itaú Unibanco

Bank

Brazil

Tribanco

Bank

Colombia

Banco AV Villas-DDDedo

Bank/Aggregator

Colombia

Banco Davivienda-Puntored

Bank/Aggregator

Colombia

Banco de Bogotá

Bank

Colombia

Bancolombia

Bank

Colombia

BBVA

Bank

Colombia

Efecty

Aggregator/Postal

Colombia

Movilred

Aggregator

Colombia

ViaBaloto

Aggregator/Lottery

Mexico

7-eleven

Retail

Mexico

BanCoppel

Bank/Retail

Mexico

Compartamos/Yastas

Bank/Microfinance

Mexico

Oxxo

Retail

Mexico

Telecomm

Aggregator/Postal

Mexico

Wal-Mart

Retail

Paraguay

Tigo Money

Mobile Network Operator

Source:  Author estimates based on latest public information available for 2015.

 

Brazil is an example of how banks can achieve proximity and simplicity through an organic growth path. The initial regulation on banking agents was issued there in 1966 and 1973. In contrast, these innovations were brought to the Paraguayan market in 2008 by a telecom newcomer, Tigo Money, in the likes of M-Pesa.

The scale of the financial infrastructure deployed by these providers is also comparable to M-Pesa. For example, Colombia has around 50,000 physical agents who cover 100% of their municipalities (see this recent CGAP blog). Brazil also has 100% municipality coverage with almost 300,000 agents, of which half process bill payments and cash-in/out (the majority of agents there manage credit products).

The volumes of transactions are comparable with M-Pesa as well. For example, Oxxo’s more than 14,000 stores in Mexico processed around 85 million financial transactions per month in 2015, including their own card-based wallet cash-in/out Saldazo, as well as bill payments and airtime top-ups. Oxxo is also the country's largest banking agent network.

This plurality of financially inclusive providers in LAC raises a question about the future of financial inclusion: How well-equipped are these providers, different to M-Pesa, to offer a richer portfolio of quality financial services to the financially excluded at reasonable prices, beyond bill payments and P2P?

Are over-the-counter transactions detrimental to offering other financial services such as insurance or savings? Is credit offered by retailer chains already acting as agents or banks an empowering tool for lower-income segments? Are these vast financial agent networks platforms conducive to making payments, including merchant payments, across the value chain more efficient?

At CGAP, these are some of the questions we are working on – questions that are relevant not only to greater financial inclusion in LAC but also in other regions in the world such as South Asia, and even for some countries in Africa, such as South Africa.

A note from the author: For the sake of brevity, I left out of the table some markets with relevant providers in terms of scale, such as Banco de Credito del Perú and Banco Estado de Chile, as well as Honduras and El Salvador where Tigo Money is also thriving. Please let me know in the comments section about any other large-scale providers I may have left out.


Lea en español / Read in Spanish on the Microfinance Gateway.

Comments

Submitted by Juan on
Dear Pablo, interesting overview. Thanks for sharing. I'm particularly interested in the transaction costs, which are of course absorbed by end users. Even though both proximity and simplicity are being achieved in countries such as Brazil and Colombia, I wonder what the efficiency is and how it would be possible to establish a benchmark. For instance, banking in Colombia is disproportionally more costly for end users than in other countries. I think that the status quo in Latam discourages innovation as sometimes it seemed as the problem was already solved.

Submitted by PGA on
Great point on the varying efficiency levels across providers: sending USD 30 to costs around 7% using Efecty in Colombia and around 4% using M-Pesa in Kenya (unregistered recipient option). Check these two links: https://www.efecty.com.co/calcula-tu-giro-lightbox http://www.safaricom.co.ke/personal/m-pesa/get-started-with-m-pesa/m-pesa-tariffs

Submitted by John BaRoss on
When discussing the difference with M-Pesa's traction between Kenya vs. India, Vodafone M-Pesa's Michael Joseph explained (at global payment industry event 'Innovation Project' held at Harvard University) a couple of key environmental similarities and differences. Both countries have in common significant populations living in rural, remote areas that are not adequately served by traditional financial service providers. A difference is that crime/theft of monies being transported was a important marketplace pain-point in Kenya but not India, suggesting a key reason for the difference in adoption of M-Pesa in these two countries. Regarding Kenya vs. South Africa, the population distribution appears to be the key factor that impacted M-Pesa's adoption. In South Africa, a vast majority of the population is concentrated in urban areas that today have an adequate presence of traditional financial service providers ... so the 'opportunity' for M-Pesa was smaller as lack of access to financial service providers in South Africa in not a market pain point of significance. Separately, it is interesting that LAC has the lowest percentage of Mobile Network Operators involved in providing, or at least involved in operating, mobile money solutions in developing nations globally. Other than MNO Tigo (noted in Paraguay, El Salvador and Guatemala), MNOs in LAC are simply used for access-only by banks and other providers of mobile money ... while across Africa, Asia, MENA and elsewhere there are many MNOs involved in operating mobile money to help optimally advance financial inclusion - faster.

Submitted by PGA on
Thank you very much for your comments. It is indeed a very dynamic environment and the future is not obvious. On risk and security I think "tiering" is one of the answers that we are seeing, check this other article: http://www.cgap.org/blog/colombia%E2%80%99s-recipe-100-agent-coverage-aggregation-sharing I also received feedback that I omitted Banco Pichincha's Mi Vecino network from Ecuador, which has around 15.000 agents. Thanks also for pointing this out.

Submitted by Steven H on
Fantastic points - particularly re: "Are over-the-counter transactions detrimental to offering other financial services such as insurance or savings?" Transaction fees are financially EXclusive. They disproportionately affect smaller transactions, which are typically conducted by those with lower incomes. In Indonesia, less than 30% of adults save money formally, yet almost 70% save in some way or another. I believe the biggest obstacle to improving savings among the poor are the huge profit pools that are transaction fees that incumbents will protect.

Submitted by Stan on
Dear Pablo, This is a very interesting piece of blog. It is important to compare how financial inclusion was led by different types of providers depending on the region and infrastructure available. Many Latin American countries have put payment laws in place to allow payment institutions (such as mobile money), in order to increase competition in the financial sector. I take Brazil as an example, where the regulation is in place since 2013, but no payment institutions were authorized so far. I wonder if these regulations will have the expected outcome. Let's wait and see.

Submitted by BarbaraMagnoni on
Brilliant analysis and thought about where this puts Latin America and all the remote models when it's time to go deeper. One thing I have been thinking is thy a lot of these models will have to tackle security as the transactions get larger and mOre frequent. The security system of a Yastas agent is not the same as a Banco do Brazil branch. Where is your thinking going on that?

Add new comment