A Digital Financial Services Revolution in Kenya

A Digital Financial Services
Revolution in Kenya:
The M-Pesa Case Study
A Digital Financial Services
Revolution in Kenya:
The M-Pesa Case Study
By
Njuguna S. Ndung’u
About African Economic Research Consortium (AERC)
Established in 1988, African Economic Research Consortium is a premier capacity

building institution in the advancement of research and training to inform economic
policies in sub-Saharan Africa. AERC’s mission rests on two premises: first, that
development is more likely to occur where there is sustained sound management of
the economy; second, that such management is more likely to happen where there is
an active, well-informed cohort of locally-based professional economists to conduct
policy-relevant research. AERC builds that cohort through a programme that has
three primary components: research, training and policy outreach. The organization
integrates high quality economic policy research, postgraduate training and policy
outreach within a vast network of researchers, universities and policy makers across
Africa and beyond.
Networking – the linking of individuals and institutions in a knowledge-sharing,

experience-sharing framework – is the key strategic instrument for implementing
AERC’s activities. The network approach links economists within and outside the
region and promotes professional esprit de corps. The Consortium is itself a network
of funders who support a commonly agreed programme of research activities, its
dissemination and the training of future potential researchers. The Board of Directors
sets broad policy, provides support for a multi-year programme of activities, approves
annual work programmes and budgets, and appoints the Consortium’s international
staff. An independent Programme Committee sets the research agenda, advises on
scientific matters and reviews and approves proposals for research and training grants.
Academic Boards for the collaborative master’s and PhD programmes oversee the
implementation of their respective programmes. A small Secretariat, based in Nairobi,
Kenya, manages the programme and provides technical support to researchers,
students and participating institutions. This organizational structure allows for
ownership of AERC activities by the network of local researchers, an independent
determination of the research agenda, and a programme of activities that is responsive
to the professional and policy needs in the region, while at the same time ensuring
accountability to funders.
A Digital Financial Services Revolution in Kenya: The M-Pesa Case Study
Published by: African Economic Research Consortium

P.O. Box 62882 City Square Nairobi 00200, Kenya
ISBN: 978 9966 61 112 3
© 2021, African Economic Research Consortium

T HE P-M ESA C ASE S YtUD
Table of Contents
List of Tables......................................................................................................................v
List of Figures...................................................................................................................vi
List of Abbreviations and Acronyms...............................................................................vii
Foreword.........................................................................................................................viii
Acknowledgements...........................................................................................................x
1. Introduction......................................................................................... 1
2. Background......................................................................................... 5
2.1 Financial Exclusion................................................................................... 5
2.2 Policy Changes......................................................................................... 7
3. Initial Experiments and Policy Frameworks............................................ 9
3.1 Initial Experiments................................................................................... 9
3.2 The Operation of M-Pesa....................................................................... 11
3.3 Risks and their Mitigation...................................................................... 12
3.3.1 CBK’s authority and resources.................................................... 12
3.3.2 General risks................................................................................. 14
3.3.3 Protection of customers’ funds through
trust account.......................................................................................... 17
3.4 Policy Considerations............................................................................ 20
4. Evolution in Use..................................................................................23
4.1 Initial Growth.......................................................................................... 23
4.2 Initial Audit............................................................................................. 27
5. Generations........................................................................................33
5.1 Generation 1: Initial Mobile Money Growth and
Financial Inclusion................................................................................. 33
5.2 Generation 2: Linked Bank Deposits and
Partnerships with Financial Institutions............................................... 44
5.2.1 M-Shwari....................................................................................... 45
5.3 Generation 3: Linked Credit Scores....................................................... 51
5.4 Generation 4: International Remittances.............................................. 53
5.5 Generation 5: Partnerships with Fintech Firms.................................... 55
5.6 Generation 6: M-Pesa Impact on Monetary
Policy and Beyond................................................................................. 58
iii
AD GIT aL F CIaLN S CEsVIR R NOUTILEV IN K ENY a
6. Lessons, Challenges and the Future for Digitization................................63

6.1 Lessons Drawn from Kenya................................................................... 63
6.2 Lessons from Use of M-Pesa as a Tool for Financial Inclusion............. 64
6.3 Replication of Regulation and other Approaches................................. 66
6.4 The Challenges....................................................................................... 67
References ..............................................................................................71
iv
List of Tables
Table 2.1: Financial access by category 2006...............................................................5
Table 4.1: Monthly transactions..................................................................................24
Table 5.1: The entry of the MFS providers in the market............................................34
Table 5.2: M-Pesa agents, accounts and transactions..............................................39
Table 5.3: Financial inclusion profile in Kenya 2006-2019.............................................
(% of adult population)..............................................................................41
Table 5.4: Creating a virtual savings base: M-Shwari................................................46
v
List of Figures
Figure 3.1: M-Pesa operations................................................................................... 12
Figure 5.1: Total customers since March 2007 (millions)......................................... 38
Figure 5.2: Value and volume of M-Pesa transations,..................................................
March 2007-August 2020......................................................................... 39
Figure 5.3: Mobile financial services agents............................................................. 40
Figure 5.4: Progress in financial access in Kenya, 2006-2019 (%)........................... 42
Figure 5.5: Financial access touch points................................................................. 42
Figure 5.6: A regional comparison............................................................................ 43
Figure 5.7: Growth in branch network in Kenya....................................................... 47
Figure 5.8: Growth in bank deposit accounts in Kenya,...............................................
2005-2019................................................................................................. 48
Figure 5.9: Agency banking uptake in Kenya, 2010-2019........................................ 49
Figure 5.10: Utilization of financial services across the providers: ...............................
2006-2019................................................................................................. 50
Figure 5.11: Adults using financial services providers (million)................................ 51
Figure 5.12: Active digital accounts in Kenya 2016 versus 2019................................ 52
Figure 5.13: Active digital accounts, country comparison......................................... 53
Figure 5.14: Migrant remittance inflows (US$ millions)............................................. 54
Figure 5.15: Cost of sending money to Kenya by corridor......................................... 55
Figure 5.16: Future strategies of banks regarding fintech............................................
companies worldwide 2017.................................................................... 57
Figure 5.17: Velocity of money..................................................................................... 58
Figure 5.18: Money Multiplier (M3/RM)....................................................................... 59
Figure 5.19: Decline in cash outside bank in relation to................................................
broad and reserve money....................................................................... 59
Figure 5.20: Increase in financial depth in Kenya (%)................................................ 60
vi
List of abbreviations and acronyms

AFI Alliance for Financial Inclusion
AML Anti-Money Laundering
ATMs Automated Teller Machines
BOT Bank of Tanzania
CA Communications Authority (CA) of Kenya
CBA Commercial Bank of Africa
CBK Central Bank of Kenya
CCK Communications Commission of Kenya
CEO Chief Executive Officer
CFT Combating Financing of Terrorism
CRBs Credit Reference Bureaus
DfID Department for International Development (UK)
DFS Digital Financial Services
DTM Deposit-Taking Microfinance
EAC East African Community
FATF Financial Action Task Force
FSD (K) Financial Sector Deepening (Kenya)
GSMA Global System for Mobile Association
KBA Kenya Bankers Association
KCB Kenya Commercial Bank
KDIC Kenya Deposit Insurance Corporation
KEPSS Kenya Electronic Payment and Settlement System
KYC Know Your Customer
MFBs Micro-Finance Banks
MFS Mobile Financial Services
MM Mobile Money
MNOs Mobile Network Operators
MSMEs Micro Small and Medium Enterprises
MVNOs Mobile Virtual Network Operators
NBFIs Non-Bank Financial Institutions
NPS National Payments System
SACCOs Savings and Credit Cooerative Societies
SIPS Systemically Important Payment System
SMS short message service
US United States
vii
Foreword
In 2007, I participated in the launch of M-Pesa where Prof. Njuguna Ndung’u, the
Governor of the Central Bank of Kenya, stated that he had allowed M-Pesa to operate,
but that the Central Bank would regulate M-Pesa once it understood what it was
regulating (this is what in later years has been described as ‘Test and Learn’ approach
and a Regulatory Sandbox). That moment represented the start of M-Pesa’s public
journey, which became proof of a concept that has influenced the development of
financial services across the world. Today, because of M-Pesa and the agent banking
that followed in response, over 86% of adult Kenyan’s are financially included. It is
an impact no one at the time could have predicted.
In this paper, Prof. Ndung’u provides a unique, personal perspective of the development
of M-Pesa. He discusses the need as a regulator to balance innovation and systemic
risk while promoting competition. Competition was of itself a difficult concept when
regulating an innovation being launched by a monopoly into a market for financial
services containing dominant institutions. He is frank about the pressure he faced
from the banking sector and policy makers and how these challenges were addressed.
The paper characterizes generations of M-Pesa, covering initial growth, bank linkages,
digital credit, international remittances and fintech. It further notes the monetary policy
impact of M-Pesa in the velocity of money circulation and the money multiplier in the
money supply process that pushed the changes in monetary policy framework. These
generations demonstrate how the continuing journey of M-Pesa, which has hugely
benefited Kenyans, has facilitated rapid growth in financial technology and stimulated
digital government and digital transformation. The story continues, and there will be
future M-Pesa generations. M-Pesa’s success is lauded both by Prof. Njuguna Ndung’u
and his successor at the Central Bank, Patrick Njoroge.
Despite this success, or arguably because of it, policy makers must now contend with
new challenges: to competition, technology-based barriers to entry, interoperability,
data protection, data privacy, cybersecurity, disintermediation, and financial
technology - in a system dominated by a few successful providers.
The final section of the paper focuses on learnings and outcomes from the Kenyan
experience. These lessons include the need to develop or enhance ‘information capital’
systems, the impact on payment systems and social payments, and the impact on
financial innovation of M-Pesa. Lessons for regulators are provided on how to promote
innovation through innovation offices, sandboxes, and reg-tech.
viii
Prof. Ndung’u closes the paper by discussing future challenges, including those
related to full account interoperability, and the need for internet connectivity across
the country for people to be able to participate in the digital finance revolution, and
electronic identification system to secure the market. He notes the challenges for
regulators to address, including emerging cybercrime, and the need to create an
enabling framework for alternative finance - specifically crowdfunding.
For me, the importance of this paper comes from the insights from Njuguna Ndung’u’s
unique perspective as a Professor of Economics and as a Governor of the Central Bank
of Kenya. He was someone who followed principles with purpose and who navigated
between multiple interests to help give birth to a transformational vehicle for financial
inclusion and the start of digital finance revolution in Kenya. Certainly, much remains
to be done, but so much has been achieved.
David Cracknell
Director, First Principles Consulting
ix
Acknowledgements
This Case Study of M-Pesa: From Trading with Pre-Paid Airtime to M-Pesa and to Digital
Financial Services Revolution in Kenya was started when I was a Visiting Fellow of
Practice at the Blavatnik School of Government, Oxford University, under the Yaw
Adjepong-Boateng Memorial Fellowship, from November 2015 to March 2016. The
case study aims to contribute to discussion around appropriate regulatory and policy
issues of mobile money by focusing on actions that the Central Bank of Kenya took
to help M-Pesa grow in a sustainable way. Being in Blavatnik was a chance also to
introduce the aspects of political economy of policy making in this case study.
I would wish to extend my deepest gratitude to Prof. Ngaire Woods, who made the
Visiting Fellow of Practice possible at the Blavatnik School of Government and working
with Prof Emily Jones who made my stay quite comfortable. This Visiting Fellow of
Practice was so paramount that a range of articles have since been published on this
subject matter. These articles include the following:
• Cashing in on digital revolution: “Digitization makes finance accessible, lowers

costs and creates opportunity”, Finance and Development, Volume 53, Number
2, June 2016;
• Practitioner’s insight: M-Pesa, a success story of digital financial inclusion,

Blavatnik School of Government, University of Oxford, July 2017;
• “Boosting transformational technology: Creating supportive environments for

game-changing innovations”, Foresight Africa: Top Priorities for 2017. Brookings
Institution, January 2017; and
• “Digital technology and state capacity in Kenya”, CGD Policy Paper 154, August
2019, available at https://www.cgdev.org/publication/digital-technology-and-
state-capacity-kenya.
In addition, I would like to express my deepest appreciation to Jonathan Espie

Greenacre, currently at the University of Boston, for having patiently edited and
commented on several sections. I am also deeply indebted to David Cracknell of First
Principles Consulting for valuable advice and writing the Foreword to this case study
and finally to Dr Anne W. Kamau of the Central Bank of Kenya, who worked tirelessly
to provide research assistance.
x
Besides these specific individuals, I acknowledge the involvement of the policy makers
who participated in shaping this research agenda. Their efforts contributed to both
the quality of the finished product, and its utility to the policy environment. Their
work stands as a valuable reference for various ministries and negotiators across the
continent. To all of you and to the many others who were involved in this project in
one way or another, I say thank you very much indeed.
Prof. Njuguna Ndung’u

Executive Director
African Economic Research Consortium (AERC)
xi
1
introduction
‘Mobile money’ is a mobile phone-based electronic funds storage and transfer service.
The whole development was enabled by the process of changing cash into electronic
units of money, e-money, and store in the mobile phone. A user can deposit, store,
transfer and withdraw funds from their mobile money account, much like a bank
deposit. The firm providing the service is normally a mobile network operator or
another type of non-bank, called a ‘mobile money firm’ (‘MM firm’) in this paper. Users
can deposit and withdraw funds through ‘cash merchants’. These are corner stores,
petrol stations, and other outlets operating on behalf of MM firms.
M-Pesa in Kenya stands out as a key so-called success story of mobile money. While
limited examples of mobile money existed in other countries, M-Pesa was the world’s
first major mobile money service. Launched in 2007, there are now 30,530,500 active
registered mobile money subscriptions in Kenya of which M-Pesa subscribers are 30.2
million while Airtel money and T-Kash are 0.344 million (Communications Authority of
Kenya - CAK, Q4 Statistical Report 2020) as at June 2020. The evolution that is critical
to its success is associated with the development of a retail electronic payments
platform that was real time and that has become a game changer in digital financial
services ecosystem.
Replicate mobile money services have been launched in other countries, creating a
global mobile money industry. According to the Global System of Mobile Associations
– (GSMA), there are now over one billion mobile money accounts in 95 developing
countries, processing a combined US$ 2 billion in transactions every day.
While growth of mobile money has been significant, it is also uneven. Mobile money
growth and use is unevenly concentrated in a small number of countries. These are
Kenya, Tanzania, Uganda and several other East African countries. Isolated examples
exist in other regions, particularly Bangladesh and Pakistan.
Regulation and policy is often given as a key reason for the uneven growth of mobile
money. Generally, Kenya, Tanzania and other countries in which mobile money has
grown tend to be classified as having ‘enabling’ regulatory frameworks. Countries
with more limited growth tend to have more prohibitive regulatory frameworks.
1
This case study aims to contribute to discussion around appropriate regulatory and
policy issues of mobile money by focusing on actions that the Central Bank of Kenya
(CBK) took to help M-Pesa grow in a sustainable way. As Governor of the CBK when
M-Pesa was first proposed and then launched, I have a range of insights into this
process. Policy makers in other countries may find these insights useful for their
own regulatory deliverables, particularly those struggling to understand how to help
mobile money grow in their jurisdictions.
The case study of Kenya is particularly useful because the CBK’s actions and those of
Safaricom, the firm providing M-Pesa, and other Kenyan policy makers, particularly
the Communication Authority of Kenya (CAK), helped spur a mobile money revolution
in the country. As at June 2020, there are now 237,637 Agents providing mobile money
accounts to 30.5 million active Kenyans, processing 4.8 million transactions valued at
Ksh 13.1 billion per day. This system relies on low-value, high volume transactions;
the average value of transactions is now Ksh 2,740.
The mobile money growth has triggered a wider financial services revolution in
Kenya. Mobile money was the first generation; others followed. Generation 2: the
Kenyan banks discovered the e-money platform as a technological instrument to
manage micro accounts and build customer deposits. Generation 3: followed quickly
naturally with commercial banks providing credit; that is loans were applied and
disbursed through the same platform. This revolutionalized the collateral technology.
Generation 4: led to M-Pesa technological platform enabling cross-border payments
and international remittances. Generation 5: has seen the collaboration between
M-Pesa and other ‘fintechs’ that have begun to partner with mobile money firms in
provision of services.
When M-Pesa was launched in 2007, financial inclusion, describing the percentage of
the populations’ access to formal financial services’, was far much lower compared
to a developed country. Just 26.7% of the population had access to formal financial
services in 2006. As at 2019, 82.9% of the population had access to formal financial
services while the excluded and informal access drastically declined to 11% and 6.1%
from 41.3% and 32.1% in 2006, respectively.
What was the role of the CBK in this process? We did many things, but our key steps
included stimulating a supportive policy environment, collaborating with other policy
makers, particularly the CAK, creating sound supervisory frameworks (here the role
of a regulator is to protect the market, develop the market, bring parties together
to design new rules that update the regulatory environment and finally nudge the
market towards the appropriate directions), and create a stable macroeconomic
environment. Of course the CBK did not take these actions at the same time. There
was an evolution in M-Pesa and wider mobile money arrangements.
2
As a whole, I can summarize five (5) major benefits that M-Pesa has made Kenya a
shining example from its evolution to the digital financial system: a retail electronic
payment system evolved, financial inclusion has been a success, sustainable business
models developed, uptake of government e-services, and improvement of tax policies.
This case study explores the CBK’s key steps in six sections. Following this introduction,
the second section provides context for the launch of M-Pesa. The third section
explains initial experiments with the service, risks, and the legal and policy evolution.
The fourth section outlines the evolution of M-Pesa based on responses from Kenyan
customers and other actors in the financial services landscape. The fifth section
provides additional data on wider developments in Kenya’s financial services system
that arose in response to the growth of M-Pesa. The sixth section concludes with
insights drawn from Kenya’s experience with M-Pesa, giving main lessons, challenges
and suggests next steps for digitization research.

3
2
background
Two main sets of factors underpinned the context for the growth of M-Pesa.
2.1 Financial exclusion
Financial exclusion, meaning lack of access to formal financial services, was a key
feature in Kenya at the time of M-Pesa’s launch. Table 2.1 below demonstrates that
only 26.7% of Kenya’s population was formally included in the economy. Financial
exclusion was particularly significant in both the rural and urban areas.
Table 2.1: Financial access by category 2006

Fiancial Access Total % Rural % Urban % Male% Female %
Category
Formal 26.7 23.8 35.5 33 21
Informal 32.1 35.5 21.6 26 38
Excluded 41.3 40.7 42.8 41 42
Source: FinAccess Survey, 2019
Financial exclusion was a particular problem for transferring money. Without access to
an electronic account, people needed to transfer money in the form of physical cash
through the use of persons, buses and other public transport vehicles and sending
friends was not considered efficient or safe. This is particularly pertinent given the
predominance of urban to rural remittances in Kenya. Those Kenyans with access to
bank accounts faced a relatively limited and inefficient banking system for transferring
money. The payments and settlement system was rudimentary and expensive, and
therefore most transactions had moved to cash and informal markets to overcome
the cost and constraints of accessing banks and other formal financial services.
There were a number of causes for high levels of financial exclusion in Kenya, revolving
around the banking system. These included the following:
• Minimum balance requirements for savings accounts;
5
• Cost of maintaining accounts – high ledger fees that reduced the amount of
funds available for savings/deposits accounts; this implies that the technology
for managing micro accounts had not yet developed;
• Restrictions on the number of withdrawals per period on a savings account;
• Low levels of income and irregular income flows that made savings in a formal
bank account not very feasible;
• Physical distance to the financial service access points; a trip to the bank either
to deposit or to withdraw was and is still expensive and adds to the barriers to
entry;
• History of weak regulatory technology and capacity that led to collapse of banks,
and leading to a dark history of institutional failure and policy failure;
• Weak legal framework and incomplete financial infrastructure – for example,

there were no deposit insurance mechanisms for many years or even information
capital to rate good and bad borrowers in banks/financial institutions;
• Customers’ information asymmetry on how banks operate/presence of segmented

markets and preference of informal market, thus customers run away from costly
formality; and
• The participants in these market segments, especially the poor and low income
earners, are sensitive to financial products and their delivery mechanism.
Lack of trust in banks and limited innovation of banks was particularly important for
financial exclusion and ultimately the launch and success of M-Pesa. In the 1980s,
the Kenyan Government sought to introduce innovation in the banking sector by
introducing non-bank financial institutions (NBFIs). These actors had lower capital
and regulatory requirements than banks and were designed to encourage such actors
to reach the poor. The problem is that these lower regulatory requirements led to
regulatory arbitrage, in turn creating weak NBFIs and their eventual collapse. The big
banks absorbed their NBFIs back into their normal operations but other stand-alone
NBFIs owned by Kenyan entrepreneurs/elites collapsed with massive savings from
poor Kenyans. The Kenyan public and business community thus held the CBK and
the government policy process with suspicion.
Other banks that focused on poor communities failed to do so. In the early part of the
2000 decade, banks that were microfinance-based in Kenya became impatient with
the Microfinance Act that was taking too long to become law. They turned into fully
fledged banks even though they were initailly deposit-taking microfinance-based.
6
2.2 Policy changes

Policy changes at various levels were underway shortly before the launch of M-Pesa.
The new Government that came to power in 2003 had the first ambition to reverse the
protracted recession that had span 1995 to 2002. It developed a blueprint for economic
recovery for wealth and employment creation, emphasized inclusive growth, but
for the financial sector. The emphasis was increasing competition, improving the
regulatory environment and reducing the burden of non-performing loans to make the
financial sector accessible. The end of this new blueprint for economic management
was succeeded by the Kenya Vision 2030. In this long-term strategic focus, Kenya was
seen as a financial hub and the financial sector reforms would be speeded up to take
advantage of the country’s location to become the Eastern African financial hub. There
were targets for savings and investment rates for the economy. But now, there were
several pieces of legislation required to push the financial sector to the next level -
the payments system, the micro finance bill, the mobile phone technology and the
regulator, the company law, the communication law and the insurance law for the
bank assurance products and also amendments to the existing law to cope with the
dynamics of market development. Given this layout, market access was important in
the banking sector; the issue was whether to push for a different market formation;
the microfinance model was on the table. The legal framework had been passed by
Parliament, but guidelines were taking too long to roll out and to be approved by
Parliament since they were considered as legal instruments. The other avenue was to
encourage the expansion of branch network of the existing banks. The discussion with
Kenya Bankers’ Association (KBA) was on how to develop a cost-effective branch outlet
structure as a delivery model for financial services to poor urban and rural locations.1
The Central Bank of Kenya (CBK) was also taking actions to develop Kenya’s financial
system, focusing on payments. In 2003, the Central Bank of Kenya Act was amended to
enhance CBK’s mandate to, “formulate and implement such policies as best promote the
establishment, regulation and supervision of efficient payment, clearing and settlement
systems.” This was therefore an additional core mandate of the Central Bank. This
mandate also required the CBK to ensure that risks associated with innovation and
technologically-driven financial services were adequately mitigated.
The Communication Bill was passed by Parliament and recognized electronic

signatures and electronic units of money. This was a great boost to the national
payments and settlement system and also to the developing concept of M-Pesa, as
discussed in the next section 2.
1 The initial discussions that I held with KBA and a few Chief Executive Officers (CEOs) of big banks
in my first few weeks as Governor was for them to develop proposals to expand branch outlets
and reduce cost outlays of brick and mortar settings model as a future proposal, but it seemed
distant to them. The cost outlay consideration was the best bet at the time but no proposals
were developed by KBA along these lines. I mentioned agency banking.
7
3
initial experiments
and policy frameworks
3.1 Initial experiments
The basic beginnings of M-Pesa can be traced back to 2002 when Kenyans started
trading with airtime. This was a simple way of buying bulk airtime as sold in shops
or in scratch cards and then slice into small units to sell or to settle debts or even to
share with friends and relatives when they ran out of airtime. In 2005, actual research
by Safaricom on the product started. But the CBK engagement with the M-Pesa
model commenced at the concept stage in 2004/5. At this time the Commercial
Bank of Africa (CBA)2, Safaricom, Vodafone, Faulu Kenya (a microfinance institution)
and MicroSave (a donor project supporting financial innovation) partnered to pilot
a payment service using mobile phones. DfID provided £ 1.0 million to develop the
product; this was a matching fund investment through the Financial Deepening
Challenge Fund, and Vodafone also invested £ 1 million.
Faulu Kenya, a credit only microfinance, conceived the idea that airtime could
be used to repay monthly instalments of loans they advanced to its customers
without a trip to the city or towns where the branches of Faulu Kenya were, for
those small repayments. The problem then was that they needed some form of
bulk or aggregators of pre-paid airtime sellers to work with Safaricom to translate
the pre-paid airtime into cash for the microfinance and effect the repayments and
servicing of the loans contracted. The initial constraint by Faulu Kenya customers
was repayment of loans by monthly instalments because it was expensive to travel
to towns where branches of Faulu Kenya were. The idea was to make these small
payments by purchasing and sending prepaid airtime. Therefore, the M-Pesa project
then revolved around sending pre-paid airtime as instalments for loan repayments
2 The bank merged with National Industrial Credit (NIC) bank effective October 1, 2019 to NCBA,
see CBK press release dated September 27, 2019, https://www.centralbank.go.ke/uploads/
press_releases/12036447_Press%20release%20-%20merger%20of%20CBA%20Limited%20
and%20NIC%20Group%20PLC_.pdf.
9
and then partner with Safaricom for aggregators of these airtime payments to be
transmitted to Faulu Kenya to effect loan repayments for their customer accounts.
The network of Safaricom Agents who were airtime sellers would be used in the
piloting and product development stage. This means that initial M-Pesa was
conceived of entirely as a way to improve the efficiency of Microfinance institutions
and extend their reach to more customers and more remote locations” (Vaughan,
Fengler and Joseph, 2012).
The research and pilot that commenced was boosted in 2006 by the passing of the
Communications Act. The pre-paid airtime model was abandoned since money could
be transformed now into electronic units of money and exchanged at par value.
Over that time, the CBA kept the CBK updated on the pilot scheme. This was
appropriate given that the CBA was regulated by the CBK. In addition, for licencing
purposes, the initial M-Pesa was a bank product for which the CBA would be partly
responsible.
In August 2006, Safaricom approached the CBK to propose a nation-wide launch

of M-Pesa. The trials had worked well and Safaricom believed there was potential
significant economic value in launching M-Pesa. Safaricom also engaged in discussions
with the Communication Commission of Kenya (CCK), and the Ministry of Information
and Communication who had a dynamic Permanent Secretary by the name of Dr
Bitange Ndemo, and the private sector players.
At the time, there was significant opposition to the launch. There was severe resistance
among members of the CBK and other commercial banks. By the time I joined the
Central Bank as Governor in March 2007, the piloting had produced some comfortable
outcomes but the real launch of M-Pesa as a bank product had not been resolved or
even discussed at the Central Bank. I was approached by different stakeholders, and
my own deputy then warned of massive bank failure if the M-Pesa was allowed into
the market. I kept an open mind and listened to all stakeholders, including those who
had no stakes, but in the process there were more positive merits of the product than
the perceived risks.
For me to proceed, I requested for a demonstration of how the product would work in
the market and all stakeholders of M-Pesa were present. I requested for forthrightness
in asking questions especially by my colleagues from the Central Bank and ministries
of Finance, and Information and Communication so that we could arrest any fears
relayed in the market about this product using mobile phone platform for financial
services.
10
The discussion below outlines the proposed operation of M-Pesa. Safaricom delivered
the outline of the scheme over 2006.
3.2 The operation of M-Pesa
Customers would buy and sell M-Pesa electronic money (e-money) from ‘Agents’
of Safaricom. M-Pesa Agents included airtime sellers, shops, petrol stations,
supermarkets, post offices, commercial banks and other financial institutions,
chemists and other retail outlets. The Agents played a crucial role in the model as
they were responsible for registering new clients, receiving deposits and making
payments from client accounts.
Deposits and withdrawals operated in the following way:
• To buy M-Pesa (deposit cash), the agent transfers e-money to the client and the
client pays the Agent;
• To sell M-Pesa (withdraw cash), the client transfers e-money to the Agent and the
Agent gives cash to the client; and
• Using the SIM Toolkit, clients could also transfer M-Pesa e-money to each other.
A short message service (SMS) from Safaricom would inform clients’ of the
transaction’s success.
A key issue was how funds from the public would be stored in the M-Pesa system.
All the e-money managed by M-Pesa was backed by real money in a trust account
held in a commercial bank (we discuss later the requirements and the choice of a
trust account). The requirements in this trust account or payments platform was
that the total balance should always match the amount in the total M-Pesa e-money
account. The account was held by a Trust Deed specifically set up to manage the
funds, which could only be used by clients of M-Pesa. Safaricom could not access or
use the funds held in the Trust Account. During customer transactions, no money
would enter or leave the M-Pesa trust account. Agents would purchase e-value by
depositing cash in the trust account. Figure 3.1 demonstrates schematically how
M-Pesa operates.
11
Figure 3.1: M-Pesa operations
Source: Adopted and modified from “Regulating Mobile Money: A function Approach, Greenacre (2018).”
Customers would use M-Pesa services to send e-money to each other, to send/deposit
e-money to their commercial bank accounts or withdraw (if they had subscribed to
that service) and to make payments for services using paybill or buy goods options.
3.3 Risks and their mitigation
Before permitting Safaricom to launch M-Pesa, the CBK carefully reviewed the original
service to identify key risks and how to respond to them. I argued at the time that
legal provisions and even legal amendments would always lag behind innovations and
dynamism in the market. It was therefore necessary to use the CBK avenue of issuing
guidelines to the market as stop-gap measures rather than stifle market innovations
and market developments. Several important issues required consideration.
3.3.1 CBK’s authority and resources
A key first step involved determining the ability of the CBK to issue guidelines on
time to regulate and re-direct the market and its developments. It was not until
2003 that an amendment to the Central Bank Act gave the CBK a discretion on
national payments to formulate and implement such policies as best promote the
establishment, regulation and supervision of efficient and effective payments,
12
clearing and settlement systems. From this amendment, an effective and efficient
payments and settlement system became vaguely an additional core mandate of the
CBK but with no legal framework to enforce it or to draw guidelines for the market.
This is the position CBK found itself in when Safaricom and Commercial Bank of
Africa (CBA) were developing M-Pesa and requesting the CBK to authorize a roll-
out of a mobile phone-enabled money transfer/remittances and payments system
that had evolved from the original concept of using pre-paid airtime service. But as
has been shown and a lot of write-ups have been devoted on this aspect, the CBK
requested Safaricom and CBA for risk mitigation in all aspects.
The CBK’s regulatory authority over payments emerged over time; for example, the
provisions of the Central Bank of Kenya Act whose mandate was expanded in 2003
to include Section 4A 1(d), which inter alia mandates the Bank to formulate such
policies as best promote the establishment, regulation and supervision of efficient and
effective payments, clearing and settlement systems. Pursuant to this mandate and
in a bid to adequately cater for the modernization of the payments system, the Bank
formulated and proposed the enactment of the National Payments System Act. While
the proposed Act will address low value retail payments systems such as M-Pesa, it will
also provide for Large Value Payment Systems for inter-bank payments such as the
Kenya Electronic Payments and Settlement System (KEPSS), which is a Systemically
Important Payment System (SIPS). Further to this mandate and to operationalize
the oversight function, the Central Bank developed an Oversight Policy Framework
document on payments system in Kenya.
The capacity and capability at the CBK to deal with these innovative demands
was crucial. A technical team from a cross-section of departments in the CBK
that covered banking, a new division of National Payments System (NPS), bank
supervision, and the legal services was formed. This technical team was supposed
to be operational and advisory. The team reviewed the application by Safaricom-
CBA and drew implications and consulted the Communications Authoity (CA). They
also did a comprehensive review of market developments and above all made sure
that the CBK did not stand in the way of innovators who would possibly bring the
much needed revolution in financial inclusion in Kenya. Therefore, from the onset,
the leanings of the CBK and indeed my own inclination were clearly in favour of this
evolution and innovation. This undoubtedly did not go down very well with some
officials of the Ministry of Finance, my own Deputy Governor and the Kenya Bankers
Association (KBA), as we will show later.
Over the course of collaboration and training, the CBK developed a clear policy
intention for considering how to address the regulation of M-Pesa. The CBK argued
that M-Pesa was a bank product that was unique in that it was a partnership between
a telco and a commercial bank. The telco (Safaricom) provided the transmission
function of funds via the mobile phone. Another was that M-Pesa would be a high
13
volume of participants and transactions, but have a low value retail payment transfer
system. Ultimately, the entire set up, in a summary, was to provide the payment
platform and also liquidity distribution among Agents that was efficient, effective
and transparent.
The CBK analyzed several basic risks to arise through M-Pesa and designed methods
of addressing them, subject to the policy intention discussed in the paragraph above.
The following section examines general risks and Section 2.3.3 focuses particularly
on protection of customers’ funds stored within M-Pesa because this topic required
particular regulatory innovation.
3.3.2 General risks
• Settlement risk – the risk that the flow of funds between transacting parties
would fail or would be delayed owing to credit, liquidity and operational risk or
use of a risky settlement medium that would not coordinate delivery and payment.
The mitigation required that Agents of Safaricom would settle through the trust
account at the Commercial Bank of Africa and since M-Pesa was a high volume low
value retail payment system, settlement in a sound commercial bank was deemed
adequate. In authorizing the M-Pesa service, and bearing in mind settlement risks,
the CBK placed maximum limits on transactions. Later and gradually, these limits
were revised upwards as confidence was boosted and payments platform was
diversified to other banks.
• Foreign exchange risk (Herstat risk) – the risk that one party to a foreign exchange
transaction would not receive the foreign currency it paid for. Safaricom had
proposed a foreign remittance service that required mitigation of this risk. The CBK
reviewed the application and advised Safaricom on the requirement of this service
in line with the provisions of the Central Bank of Kenya Act.
• Legal risk – the risk that unexpected interpretation of the law or legal uncertainty
would leave the payment system or members with unforeseen financial exposure
and possible losses. In mitigating this risk, the CBK considered that Safaricom as a
mobile service provider was licensed and regulated by the then Communications
Commission of Kenya (CCK) and was therefore under the provisions of the Kenya
Communications Act 1998 and the amendments in 2006. The legal relationships
with respect to money transfer are provided for in various agreements that were
reviewed by the Central Bank of Kenya. In addition, there was recognition that this
innovative service would be a value added on the licensed mobile services and as
a money transfer service, it would benefit from the existing oversight mandate of
the Central Bank of Kenya. This would further be enhanced by the enactment of
the proposed National Payments System Bill.
14
• The Anti-Money Laundering (AML) and Combating Financing of Terrorism (CFT)

– from the outset and even before authorization, Safaricom was advised on the
requirements of AML/CFT. In this respect therefore, Safaricom continued to mitigate
this risk through legal instruments, training its Agents, monitoring the system for
suspicious transactions, and enforcement. In addition, Safaricom observed AML/
CFT policy requirements of Vodafone. Just like any other institution in Kenya then,
Safaricom would await the enactment of the AML Bill, the terrorism financing law
and the institutions to safeguard and protect the market, the Financial Intelligence
Unit (or Financial Reporting Centre as Kenyan Law later defined it), where all reports
on suspicious transactions would be reported and actions taken.
• Operational risk – the risk that hardware or software problems, or human error, or
malicious attack would cause the system to break down or malfunction, giving rise to
financial exposures and possible losses. The comfort was that Safaricom being part
of the Vodafone group, an international and reputable multinational in the provision
of mobile technology, the M-Pesa product would continue to benefit from research
and development of Vodafone. Operational issues were regulated by the CCK, now
CA. The CBK would continue to receive reports on operational issues on a monthly
basis. The CBK also emphasized the need for Safaricom to ensure adequate disaster
recovery and business continuity arrangements.
• Systemic risk – this was a payment solution fully backed and comprised a small
proportion of the payment system. To date, the mobile phone payments in total
only comprise 5.5% of total national payments. Therefore, it was not significant
and since it was fully backed up it did not pose any systemic risk in the payments
system. Systemic importance requires particular consideration. Even at present,
the risks posed by M-Pesa continue to be within levels considered not systemic.
Though it has a high volume of participants and transactions, M-Pesa is a low value
retail payment transfer system. The CBK continued to engage Safaricom around risk
mitigation measures to ensure the continued safety and efficiency of the service.
The measures were then set:
- Limiting the size (value) of the mobile transaction; this was set at Ksh 35,000
(US$ 350) per transaction at any one time.
- The SIM card could not hold more than Ksh 50,000 (US$ 500) at any one time.
This maximum limit would discourage the SIM card holders from making it
look like an alternative bank or holding account.
- There were daily limits on transactions and only two transactions were allowed
at the limit of Ksh 35,000 per day. This made sense for monitoring AML/CFT
regime in Kenya. During this time, even with all these developments and
innovations taking place in Kenya, the AML/CFT regime in Kenya as per the
15
Financial Action Task Force (FATF) classification was considered inadequate (in
subsequent years Kenya was placed under the ‘dark grey’ list with possibilities
of counter measures). For this reason, stringent measures were put in place
at the time to adhere to AML/CFT requirements at the bank and at the M-Pesa
platform levels3.

However, the thresholds limits were increased later on due to market pressure as
more and more joined the payments of goods and services categories. At first, the
Central Bank escaped the pressure of raising the limit by arguing that Safaricom
should increase the number of Agents since transactions value were still low but
later, as transactions value increased due to entry of utilities that boosted the
functionality of payments for goods and services, the limits were increased and
in the subsequent years 2010-2011 a super user category (and also the original
concept of aggregators) with no limits was introduced for specific Agents, bank
tellers and supermarkets point of sales. But in addition, the technology has
supported a better and more efficient AML/CFT monitoring framework and also
the Know Your Customer (KYC) tiered approach has supported the CBK’s risk-based
approach appropriately. As it stood, M-Pesa had accomplished its objectives and
reduced the fear in the market of systemic risk and that perhaps was going to drive
further innovations in the market.
Protection of customers’ funds with cash merchants required particular attention

from the CBK. Two risks were particularly important.
- Credit risk – the risk that a counterparty may fail to meet an obligation for
full value either when due or at any time thereafter.
- Liquidity risk – the risk that a counterparty would not settle an obligation
for full value when due, but at some time thereafter.
In consultation with the CBK, the M-Pesa model addressed credit and liquidity risks
through Agents in a number of ways. To address credit risk, Agents would prepay before
offering services to the customers. The Agents would sign to agreements enforced by
Safaricom and also agreed with the CA and for which the CBK had an input.
3 I recall in 2014 FATF Plenary in Paris I argued in defence for Kenya financial system that had been
placed in ‘Dark Grey List’ for countries with inadequate AML/CFT regimes, that M-Pesa platform
had allowed informal transactions to move to formality and that they could be monitored. In
my argument then, M-Pesa was fighting informality of financial market transactions and that
informality was more dangerous to the AML/CFT regime for Kenya. The view held about M-Pesa
was completely the opposite. My view was well taken. I argued that Kenya and indeed the African
continent, the informality of markets and financial exclusion was more dangerous for AML/CFT
regime. This point was appreciated and explains why FATF criteria has now pushed for financial
inclusion agenda to improve AML/CFT regimes.
16
Safaricom addressed liquidity risk at Agents by employing stringent vetting criteria

that included entities that were financially sound. As at September 2008, the Agents
were distributed as follows: Commercial banks (3.9%), Postal Corporation (2.2%),
Supermarkets (1.5%), Petrol stations (1.3%), Shops (24.5%), Airtime sellers (60.7%),
Savings and Credit Cooperative Societies - (SACCOs) (3.2%), Courier companies (0.9%),
and Chemist shops (0.9%). Subsequently, and following discussions with CBK, other
participants such as commercial banks and Automated Teller Machines (ATMs) were
included.
Other regulatory developments were also relevant. For example, in 2006 the
Communications Bill recognized in law the electronic signature and the electronic units
of money. Therefore, a new product that would turn cash into electronic units of cash
was now legitimate and therefore the mobile phone SIM cards would store value. The
old concept of pre-paid airtime was thus modified to fit with these new developments.
From a market reception point of view, those who participated in trading with pre-paid
airtime became the ready market for early adapters of this new product M-Pesa. Perhaps
this has been a challenge in other countries because the concept of Agents network
exchanging cash for electronic units of cash was not well understood.
3.3.3 Protection of customers’ funds through trust account
The CBK’s classification of M-Pesa as a banking product is particularly important

for protection of customers’ funds. The CBK and CA agreed on the basic regulatory
framework on this understanding. In most countries that seem to have failed to take
off in Digital Financial Services (DFS), they have described the Kenyan model as ‘Telco-
led model’. However, they have failed to see through it that if telco did not partner
with the banks, then this model would not have worked. But more importantly, at
the end of the process, M-Pesa was a bank product and the telco was a partner in the
investment and was also providing an efficient transmission system. This point has
been overlooked and therefore quite misunderstood.
Regulatory innovation was also needed in regulation because there was no national
payments and settlement law in Kenya. This was a particularly thorny issue in Kenya
and in 2009, the Central Bank of Kenya through the support of Financial Sector
Deepening Kenya and the Bill and Melinda Gates Foundation contracted Bankable
Frontiers to develop the National Payments System (NPS) draft guidelines. We called
them ‘draft guidelines’ because the NPS Act had not been passed by Parliament as
a legal framework to back the guidelines. But the CBK used the ‘draft guidelines’ to
regulate and protect the market.
The idea was that the M-Pesa trust account would be efficient with several layers of
insurance and the trust law would be adequate to protect the payments platform. It
would be viewed as a deposit account in the commercial bank where the Central Bank
17
had adequate regulatory capacity and control over commercial banks. But also, telcos
would have no access to the funds, they were just transmitting the funds to the trust
account. The CBK would prescribe from time to time guidelines on residual balances
that would earn interest but now under the Trustees. Initially, it was guided that such
funds could only be withdrawn and used for corporate social responsibility (CSR).
Safaricom’s Agents would deposit money in a trust account managed under trust law
in Commercial Bank of Africa. Safaricom could not access these deposits, which were
ring-fenced under the trust deed and as per guidelines the CBK prescribed and were
later adopted by the National Payments System Regulations. The money in this trust
account is not under the control of Safaricom and cannot be employed for purposes
such as lending, investing or in any other manner for the account and at the risk of
Safaricom as per Section 2(1) of the Banking Act. This is like reverse engineering - the
guidelines were based on the workings and experiences of M-Pesa and they would
now come back not only to influence the regulatory process but also the DFS policy.
The Trust Account addressed the credit and liquitidy risks in the following manner:
1. Credit risk
The model allowed Safaricom to issue electronic money in exchange of cash at par
value and was stored in the SIM card for the customer. Once the electronic money
was stored in the SIM card, it was simultaneously loaded into the trust account at the
CBA and this account was under the custody of trustees. The rules around the trust
account is provided in the trust deed. Various legal instruments pertaining to this
service, including the trustee deed, were presented to the Central Bank for review
prior to the launch of this product. Further to this, funds in the trust account deposited
in the Commercial Bank of Africa are regulated by the Central Bank of Kenya under
the Banking Act. In the description, Trust Deed “Between M-Pesa holding company
(trustee) and M-Pesa participants” details the aspects of this service where the Trustee
holds funds on behalf of all M-Pesa System Participants under a Declaration of Trust
(the Trust Deed). Highlights of the Trust Deed are:
• The Trustee holds all amounts which constitute the Trust Fund on trust for the
System Participants.
• The beneficial entitlement of each System Participant to the Trust Fund at any
time shall be to such amount of the Trust Fund in conventional money as is equal
to the amount of E-Money in the M-Pesa Account of such System Participant at
such time.
• Safaricom is entitled to levy certain charges on System Participants for the

operation of the service. Where it does so, the M-Pesa Account of the relevant
18
System Participant will be debited by the amount in E-Money of the relevant charge
and a M-Pesa Account of Safaricom shall be credited with the relevant amount.
• The amounts constituting the Trust Fund shall be held by the Trustee in Commercial
Bank of Africa.
• Safaricom undertakes to the Trustee and to the System Participants that it will not
issue any new E-Money other than in return for an equal amount in conventional
money being paid to and received by the Trustee.
• Safaricom shall also not effect any transfer of any E-Money from any M-Pesa
account of an amount which exceeds the credit balance of E-Money in the relevant
M-Pesa account.
2. Liquidity risk
Trust account addressed liquidity risk by storing funds (trust assets) in a bank account.
Assuming the liquidity and solvency of banks, customers’ funds should always be
available – in liquid form – to address liquidity problems that might arise through
the M-Pesa service.
The basic M-Pesa trust arrangements have been used widely. For example, even when
the NPS Act was finally passed and the CBK developed and rolled out the guidelines,
we still find to date new products coming into the market partnering with banks and
opening trust accounts as the preferred payment solution platform. Perhaps it is still
seen as the most robust platform to support payments solution in Kenya. This means
that it was still the best platform for e-money distribution because it allows Agents in
all corners of the country tied to a payments platform to provide financial services with
ease. The conclusion to make then is that Safaricom and CBA supported by the Central
Bank of Kenya generated ideas and developed a versatile payments platform that has
stood the test of time. Looking at literature and subsequent research that has taken place
since its launch, this platform adequately demonstrates that transactions can take place
effectively, efficiently, in real time and can be easily traceable and monitored (by the
regulator for systemic concerns, safety and soundness and also for AML/CFT purposes).
Safaricom supervised and regulated its Agents. It has been a major lesson for the
network of Agents who formed the point of service countrywide. The Agents’ network
by Safaricom formed the backbone of investment for M-Pesa to work and serve the
Kenyan population. The Central Bank of Kenya regulated commercial banks and
so the trust accounts. Therefore, once money is stored in the SIM card and also
simultaneously in the trust account, only the account holder can access the money,
just like a normal deposit account in a bank – in this case what changed is the use
of the mobile phone rather than physically walking to the banking hall and filling
19
request forms to the teller to transfer or pay for goods and services. The M-Pesa Agents’
network has dynamically developed to form a formidable set of financial access touch
points with different layers of responsibility and complexity to distribute e-money
liquidity effectively, efficiently and transparently.
CBK authorized other Mobile Network Operators - MNOs (Airtel, Orange and Yu) and
Payment Service Providers (Tangaza and Mobikash) followed soon and launched their
products. But they followed a similar model that Safaricom had created with CBA; it
became the norm. But now, more policy issues began to emerge, as the joining telcos
needed to create their own Agent networks and partner with other banks. This of course
is where the initial investment and capacity seems to matter. The banks with partnering
arrangements negotiated to link their customers’ bank accounts with the mobile phone
and to M-Pesa. At the end, first generation digital financial services (DFS) banks had
linked up with mobile phone users and the mobile phone money transfer ecosystem.
In the meantime, other telcos wanted a share in the market and therefore required the
regulations and the market environment to be levelled for their easy entry.
3.4 Policy considerations
There were several policy challenges that the Central Bank team faced:
• The first policy challenge was conformity; that is, to make sure the proposed
M-Pesa product was not, by legal framework, defined as a banking business and
deposit taking. This in a sense divided the process into two where one dealt with
M-Pesa Agents exchanging cash into electronic units of money, which was not
banking business. The second process was the transmission of those funds to CBA
Trust account that was a deposit account as well as a payment platform which
was banking business with the regulatory mandates of CBK. The Agents showed
and advised that in the first process, the design of M-Pesa did not fall within this
legal definition of deposit taking or a place of banking business. For this to be full
proof and conform to the definition, the CBK had to make sure that M-Pesa funds
were not being intermediated or accessed by Safaricom, but also demonstrate
that the transactions between the Safaricom Agent and an M-Pesa customer was
just like a shopkeeper – only that the Agent was exchanging electronic units of
money with cash while the shopkeeper exchanges physical goods for cash but it
was the same philosophy. This distinction has been quite important to date both
in Kenya and outside, and mostly in those countries that have not managed to
replicate the M-Pesa model. But in Kenya, what perhaps makes the difference is
that the law recognized electronic units of money. This of course made it easier
for the CBK to define the banking business and deposit taking more appropriately
in terms of what M-Pesa model was capable of doing and therefore gave M-Pesa
product a good start. In addition, the telcos and Safaricom were not in contact
with these funds at all and therefore were not in the banking business.
20
In the meantime, telco-agency model required further clarification and restrictions.

There were four aspects of the CBK’s legal opinion relating to the Safaricom Agents’
operations:
1. At the point of conversion from cash (bank notes and coins) to electronic
units of money, the exchange was at par value – that is the electronic units
of money was equivalent to the cash provided.
2. The electronic value of the cash exchanged would be reflected in the mobile
phone account.
3. This amount in the mobile phone account could only be accessed by the
account holder and would remain under the control of the mobile phone
account holder.
4. The conversion from cash to electronic units of cash was not under any terms,
with the Safaricom Agents taking cash as would be the case in a bank account
where a deposit is made with specific defined terms that have legal obligations.
• The second policy question was more operational and related to safeguarding
M-Pesa product and the M-Pesa model from myriad of perceived risks (see section
2.3.2 for the risks). The CBK had to make sure it was a low risk money transfer
system and these risks would not in any way be systemic then and in the future.
The reaction by the CBK was to list quantum and other measures that would even
lower the risk profile even further as already decribed in section 3.3.2.
• Finally, the CBK required the CBA and Safaricom to provide monthly returns that
would allow the CBK to monitor average transactions, average residual balances
and inactive accounts.
Surprisingly, after the demonstrations and technical details of how the product
worked, and once we resolved our plans, there were no substantial questions beyond
the rumours in the streets. We therefore agreed that all risks, due diligence and
supervisory/regulatory issues had been adequately addressed at the piloting stage
to allow a launch and continuous monitoring. I boldly blessed the project and in
doing that the CBK approved M-Pesa to be launched in 2007. As I made my remarks
in that meeting, I strongly emphasized that this would form a technological platform
for a menu of financial services that would solve the access problem in the financial
market – that has to date not been disapproved4 and has formed the first phase of DFS.
4 These issues will be emphasized later since the regulatory side was satisfied but some senior
Treasury officials and the big banks under the cover of the Kenya Bankers Association (KBA) were
quite uncomfortable and they would wage a war on CBK and M-Pesa much later.
21
4
evolution in use
This section describes the initial growth of M-Pesa, response from banks, and an initial
review of the system.
4.1 Initial growth
The picture that was emerging was an imposing network of Safaricom Agents, the
imposing green colour of Safaricom and heavy advertisement on ‘send money home
by M-Pesa’ and no one mentioned the CBA at all. The words ‘Ponzi game’ and ‘Pyramid
scheme’ were used mostly by those who had not entered or used the service. But
behind the scenes, the CBK, Safaricom and the CBA (at least) knew how the customer
funds were protected, ring-fenced and how the model worked.
The reception by Kenyans, given the confidence and the heavy advertisement by
Safaricom, witnessed a tremendous growth not only on customer base but also
increasing transactions.
What seems to explain this phenomenon growth and capacity of M-Pesa? In the
first 12 months of M-Pesa operations, the number of Agents increased from 307 in
March 2007 to 2,329 in March 2008. This is more than two times the number of bank
branches in Kenya. By August 2008, the number of Agents had increased to 3,761
(Table 4.1). The most striking is the number of transactions and the average value
of transactions from the very beginning. By the end of March 2007, the month of
the launch, there were 21,714 transactions and by April, the number of transactions
had more than tripled to 69,740. By December 2007, the monthly transactions were
over 1.2 million and by August 2008, after one year of M-Pesa operation, the total
monthly transactions stood at 6.3 million valued at Ksh 16.8 billion (close to US$ 25
million) but the average transactions value still stood at a low of Ksh 2,916 (close to
US$ 40 then) per transaction. Even though individual transactions were low value,
the number of accounts increased and the number of transactions increased in
volume. In addition, the number of accounts and the number of Agents providing
the services increased.
23
To improve on management of transactions, speed and risk mitigation, Safaricom

constantly reviewed the technological capacity and capability while at the same
time the CBK encouraged the diversification of the trust account to other banks. But
before we go into this direction that shows how the model of M-Pesa dynamically
improved its base, it is important to point out that this was not without challenges.
The CBK had to deal with various complaints from members of the public and Kenyan
parliamentarians. Most of these complaints centred on speed on M-Pesa transactions
(especially on Fridays when traffic was highest) and continuous constraints on liquidity
or float with Safaricom Agents and finally the limit of Ksh 35,000 per transaction,
which was considered too constraining. My position and response was that increasing
the number of Agents was necessary rather than adjust the limits without other
safeguards to protect the market, and that Safaricom was continuously improving
on the payments solution capacity.
Table 4.1: Monthly transactions

Total value of transactions Total number of transactions/ Average value per
(Ksh million) Accounts Agents transaction (Ksh)
Mar 07 64 0.02 21,714 307 2,965
Apr 07 221 0.05 69,740 362 3,167
May 07 484 0.11 149,986 447 3,225
Jun 07 720 0.18 233,661 527 3,082
Jul 07 1,065 0.27 354,298 681 3,007
Aug 07 1,580 0.43 516,239 819 3,060
Sep 07 2,070 0.64 669,689 960 3,091
Oct 07 2,830 0.85 958,908 1,196 2,951
Nov 07 3,515 1.13 1,221,742 1,379 2,877
Dec 07 3,770 1.35 1,274,098 1,582 2,959
Jan 08 4,059 1.59 1,346,827 1,812 3,014
Feb 08 5,220 1.82 1,739,903 2,067 3,000
Mar 08 6,747 2.08 2,397,498 2,329 2,814
Apr 08 8,390 2.37 3,072,888 2,606 2,730
May 08 10,904 2.72 4,021,265 2,770 2,712
Jun 08 10,917 3.04 4,201,440 3,011 2,598
Jul 08 14,017 3.37 5,381,073 3,378 2,600
Aug 08 16,756 3.73 6,342,413 3,761 2,642
Average transfer amount per transactions 2,916
Source: Central Bank of Kenya
24
It was necessary to enlarge the network of Agents as this later became a useful network
of financial access touch points and also significantly increased employment of the
youth – more versatile with the mobile phone operations in addition. The constraints
on “float” or liquidity distribution was being handled by Safaricom by introducing
aggregators and super user category of Agents. As pressure mounted to revise the
transactions’ limit upwards, two fronts were introduced; first, Safaricom improved
the tracking and monitoring capacity of the system and then launched the super user
category of Agents. These were also called ‘aggregators’ who would help in e-money
liquidity distribution without limits for themselves. Second, the limits were revised
upwards to Ksh 70,000 (about US$ 700) and to hold up Ksh 100,000 (about US$ 1,000)
in the SIM card5. These two introductions to the e-money services in addition to the
rising number allowed more dynamism, more efficiency and reach for the M-Pesa
products in the market and reduced any further pressure on the limits set and the
speed of transactions.
Coming back to the model of M-Pesa, it has now been described with technical
details, and the easiest way to understand it is to relate it to where it started with
an innovative idea of using pre-paid airtime distribution model. This was the base
of cash in, cash out, airtime top-up provided by a network of Agents supervised by
Safaricom on guidelines agreed with the CA and for which the CBK had input. The
Agents registered new customers on behalf of Safaricom, and they provided the
cash-in, cash-out on the basis of allocated ‘float’ (e-money liquidity), thus they were
also Agents of distributing liquidity in the network. But in rural areas and poor urban
locations, the Agents were the source of information and operational education for
this new model and the mobile phone use – they were the face of Safaricom. This
setting is what contributed to the success of M-Pesa and allowed an easy take off.
The network of Agents in urban centres, urban periphery and rural shopping centres
were distinct in colour and very accessible. The credibility of Safaricom Agents,
their visibility, the investment by Safaricom in terms of equipment, training and
supervision gave this model the success it has to date. Of course at present and
after seeing the success of mobile phone-based financial services and a range of
products coming into the market, important issues and institutional frameworks
have emerged such as competition issues and safeguards, market dominance,
monopoly power, consumer protection and interoperability.
But behind these issues and debate is the illusion that telcos are making massive
economic rents due to mobile phone-based money transfer and other products using
the platform. The factual position is that this service makes a very small proportion
5 The limits as at October 2020 were Ksh 300,000 (US$ 3,000) in deposits, maximum daily
transactions value of Ksh 300,000 (US$ 3,000). Maximum amount per transaction is Ksh 150,000
(US$ 1,500).
25
of their profits and revenue; actually revenue from voice dominates. The crucial point
here is that M-Pesa initially did not have to make money – its value proposition was
significantly to reduce churn to competition.
But this does not mean that competition and interoperability are not important
and helpful in developing the market further. I have argued later in this paper that
interoperability will help increase and grow the market size and individual telcos will
capture the market share by the products they roll out into the market. This is because
there appears to be a misconception that interoperability will help the late starters
capture their rightful share of the market, but the reality is that they will have to roll
out their products, their Agents network (or share existing Agents), train the Agents
and enter into remote areas still left behind by inadequate networks but also invest
in real time payments. Investments in real time payments across mobile networks is
what has held back interoperability in Kenya.
Perhaps one can look at the network of Safaricom Agents as an infrastructure to

provide capacity for growth in this service as it was planned then, and how it has
worked since, including taking advantage of the supporting policies rolled out by the
CBK, such as Agency banking model. First, it was a massive investment to build and
maintain such a network of Agents with equipment, training and daily surveillance.
Perhaps this may explain why other mobile phone operators entering this market have
found it difficult to match the investments, but these investments have to be made
for meaningful competition to take place in Kenya. In addition, I do believe there was
an incomplete strategic grasp of mobile money market by the competitors wishing
to enter the Kenyan market and how it was also reflected by the scale efficiencies
of Safaricom – it had many more agents, more customers and a complete physical
network extesivley expanding to cover the country.
In basic models of competition, in the first stage the firms invest in capacity. The
capacity invested determines not only its capability but also signals the presence in
the market; this is the M-Pesa case. It is important to capture the role of Agents at
this stage, since the debate in recent years in Kenya has focussed on removing the
exclusivity clauses of Agents rather than the functionality that gave the mobile phone
financial services ecosystem the vibrancy that was witnessed. As already highlighted in
section two, the role of Agents not only included exchange of e-money and provision
of services in the rural areas but also to: register new customers while observing
the KYC guidelines. The basic registration requirements were an identity card and
address, among other requirements. It is important to recall that when mobile phones
penetrated in Kenya, there were no legal requirements to register a SIM card, except
if one wanted to open a post-paid account. This also explains why the majority of
Kenyans joined the pre-paid airtime service. When M-Pesa started, the KYC guidelines
had to apply; this is part of the risk based approach by CBK. On this basis, KYC, the
tiered system was then applicable and consistent with AML/CFT regime.
26
Therefore, the totality of the Agents’ network for all the mobile network operators
(MNOs) in Kenya has played a vital role to the success of the DFS and the M-Pesa model
and the debate can be advanced along the following lines:
a) A network of financial services providers and points of service outside the

commercial banking or microfinance set up.
b) E-money transfers, payments and settlement network in real time that is

transparent, effective and efficient.
c) The DFS ecosystem can then be used for a variety of purposes given the capacity
of the Agents’ network.
d) The number of M-Pesa Agents in Kenya is so far above 200,000 and still growing
– beyond any branch outlet of banks that can be thought feasible.
With the success of M-Pesa, it meant that it was a vibrant market ready to be discovered
by other MNOs and banks wishing to enter the market. Two outcomes are worth
noting at this stage, the CBK had set a precedent for other operators wishing to enter
the market. The commercial banks were feeling threatened by this market and the
M-Pesa model adopted and taking root in the Kenyan economy. In addition, some
banks considered the investment in a payments platform was rather prohibitive for
them. This was the political economy at play; the commercial banks started hitting
at the CBK using the Kenya Bankers Association and the Acting Minister for Finance
in late 2008. Before I show the intensity of this threat and the outcome, it is important
to make two points to introduce the power play at hand. First, since the process of
launching M-Pesa had worked (it was launched by the then Minister for Finance, Hon.
Amos Kimunya) and so far the results had vindicated the expectations of the CBK as
well as those of Safaricom and CBA, it was now a fertile ground for other MNOs to try
their luck. The CBK had now to purify the framework by issuing a letter of no objection
to Safaricom to roll out M-Pesa as a stopgap measure so that it did not appear to stifle
innovation. To purify the framework, what was required was a legal framework, a
parliamentary registration of the NPS Act and in its absence a set of formal guidelines
the CBK could use as mandated by law.
4.2 Initial audit
The CBK had underestimated the threat of M-Pesa on conventional banking system.
The concern from the commercial banks was that adoption of mobile money would
lead to a “disintermediation and drive down the use of savings accounts in traditional
banks” (see Muthiora, 2015). The CBK argued that integration with M-Pesa would
support further reach to customers, increase the level of deposits and savings and
27
therefore raise the scale and intensity of the intermediation process.6 In my own
words then and in support of M-Pesa, I argued that mobile phone financial services
would form a technological platform that would allow banks to roll out a menu
of financial products and services that would reach a mass market in Kenya. In
addition, this platform would support an efficient system of transactions, managing
savings and deposit accounts and disbursing credit. This connection was not seen
as feasible by most of the commercial banks. While the tension continued behind
the scenes unknown to the CBK, and spearheaded by the KBA (a good cover for big
and multinational banks), the then Minister for Finance Hon. Amos Kimunya who
had launched M-Pesa in March 2007 resigned and was replaced by Mr John Michuki
in an acting capacity towards the end of 2008. The KBA, it appears, had a listening
and sympathetic ear.
One morning, I flanked the new Acting Minister in a press briefing after questions
arose on economic growth in Kenya, the global financial crisis and the effect on the
Kenyan economy, inflation and exchange rate. The Acting Minister did not dwell on
the responses to these questions; he turned to me and directed me and the National
Treasury to conduct an audit on M-Pesa because the information he had gathered
was that M-Pesa would not end well in Kenya. This was hardly encouraging for a
Minister of Finance either to the market or to the Central Bank. This caught the CBK
and Safaricom by surprise. But it was not a written directive, therefore the minister
was playing gallery with the media. The Minister knew that by law, he could not direct
the CBK on its operations or mandate. But the law and gallery were two different
aspects – Kenya was now at a standstill. I called Michael Joseph and my colleagues
at the CBK and made a decision to deal with the issue head-on. What the CBK and
Safaricom developed overnight was not an audit but a report to vindicate the success,
the risk mitigation and the opportunities of M-Pesa. This was shared with the National
Treasury and they were happy with it and they were comfortable to share with the
Minister. I followed this report with a draft press release that the CBK would issue to
calm and create confidence in the market once the Minister was satisfied with the
‘audit’ report on M-Pesa.
The Minister must have been impressed by the clarity and correctness of the picture
painted by the ‘audit’ report and the press brief and perhaps the power to calm the
market and also create confidence in this product. He called me to his residence for
a discussion and informed me that he was satisfied that the ‘audit’ was an accurate
reflection of what was happening in the Kenyan financial market and that M-Pesa
was the best product in our financial market and for financial inclusion. He further
6 Data at the time showed that the banking sector had only 4.3 million deposit/savings accounts,
only about 700,000 accounts for loans and advances and only 740 bank branches country wide for
an adult population close to 30 million in Kenya. This shows that accessibility was a constraint.
28
indicated that he would be the one to sign the press release7. It was a relief to me
and Safaricom, and so it mattered little who would sign the press release. I removed
my name from the press release and substituted with his, and then it was signed and
released to the market. I made sure it reached the KBA and the media as soon as it
was signed. The press brief is reproduced here and the detailed audit report in Annex
1 that accurately describes the process and the safeguards including data points of
residual balances in the trust account. The press release was a summary of the ‘audit’
report that made the following factual points:
• M-Pesa was not a banking service, it was a low value retail money transfer and
payments system.
• The CBK had approved M-Pesa as a bank product on the basis of its adequacy, its
legality, and satisfactory risk mitigation.
• It was not a deposit-taking service.
• Its funds were secure in a trust account and not accessible to any other entity.
• The CBK had a close oversight on the product and that it was emerging as a strong
instrument to effect financial inclusion in Kenya.
The headlines that followed from the media; “the regulator (CBK) gives M-Pesa a clean
bill of health” was a paradox since the CBK never doubted its capacity, operations and
even its correctness and effectiveness. This turn of events, perhaps by accident rather
than design, gave a resounding confidence to M-Pesa in the market and in doing so
other MNOs were confident to succeed. The commercial banks that had spearheaded
the war had no option but to integrate with M-Pesa and other similar products being
rolled into the market. It is surprising that the Standard Chartered Bank moved fast
to negotiate with the Safaricom to host the payments solution platform after this
confidence boost. It became the second bank to partner with Safaricom to establish
and diversify the Trust account as a payments solution and therefore drive the M-Pesa
product in the market. Henceforth, other banks negotiated with Safaricom to link
their account holders with M-Pesa, whereby they could draw down from their bank
accounts to M-Pesa accounts without having to go to M-Pesa Agents. Slowly, even
the microfinance institutions and SACCOs joined in. M-Pesa became the coordinating
payments solution across all market segments. The press release is reproduced here:
7 I did not know at the time that other concerned Kenyans had tried to intervene on this issue
and had called on the Minister. One documented aspect is where Michael Joseph, the CEO of
Safaricom, paid a courtesy visit to the Minister and demonstrated to him how to send money to
his farm manager in his rural constituency. He was impressed on how the M-Pesa worked (see
Muthiora, 2015)
29
Box 4.1: PRESS RELEASE M-PESA MONEY TRANSFER SERVICES
The Government initiatives to promote Information Communication Technology and

a fully market driven economy with e-commerce has enabled the country make great
strides in the area of financial services delivery. In tandem with these advancements,
many businesses have tended to leverage on the same and provide value adding services
that are not only innovative but have grown to become popular with the Kenyan public.
One such innovation is in the area of Money transfer. Kenya is now a world pioneer in
the use of mobile phones to transfer Money, following the rollout especially of M-Pesa
in March 2007. However, the adoption and growth of M-Pesa services has not only
continued to draw public attention but has also generated a lot of debate as to the safety
of these kinds of payment and transfer systems. It is necessary then to provide an audit
of the system to guarantee comfort on its safety, information about its effectiveness as
well as to satisfy the operating platform for M-Pesa and other similar services wishing
to enter the market.
The purpose of this press release is to provide insights as to how this innovative money
transfer service has developed, how it has enabled the transfer of funds to the unbanked
and how the Central Bank of Kenya (CBK) continues to oversee its operations in order
to ensure their safety and efficiency.
1) At the onset, I wish to reiterate that the Central Bank of Kenya and the Treasury
are committed to promoting safe and efficient innovations that enhance access to
financial services thereby addressing the challenge of financial exclusion occasioned
by infrastructural constraints to formal banking services. Since inception, the number
of registered M-Pesa users has grown to about Four (4) million. These users are
served by M-Pesa Agents that are spread across the country including remote rural
or poor urban areas who previously did not have access to formal banking services.
The service has therefore proved to be an effective way of reaching the unbanked
members of the Kenyan society. The popularity of the system stems from its ability
to transfer values at more affordable rates to the public. So far, the system has
maintained an average Ksh 3000 per transfer. In the past, commercial banks have
found it not efficient or effective to offer such services. However, it is laudable to
note that some commercial banks and other service providers are now partnering
with M-Pesa with a view to complementing each other and leveraging on the M-Pesa
outreach.
2) Prior to the launch of M-Pesa services in Kenya, Safaricom sought authorization from
CBK to undertake the money transfer service. In evaluating the proposal, the Bank
considered the request on the basis of the safety and efficiency of the services. In
addition, precautionary measures were put in place to ensure that the services did
not infringe upon the banking services regulatory framework as provided for under
section 2(1) of the Banking Act. The service therefore does not:
a) Accept from members of the public money or deposit that is repayable on

demand or at the expiry of a fixed period or after notice;
30
b) Accept from members of the public money for current account purposes that
is used for payment and acceptance of cheques; and;
c) Employ money held or any part of the money for purposes of lending and
investment or in any other manner for the account and at the risk of the person
so employing the money.
3) The Central Bank has further confirmed that funds to be transferred are held in a trust
account in a commercial bank account and is therefore not available to Safaricom for
lending, investing or operating as an ordinary bank account as described in section
2(1) of the Banking Act. These funds are held in trust for the benefit of customers. The
trust deed provides legal protection for the money in the trust account and provides
mitigation against the potential risks customers may be exposed to. Accordingly,
compliance with this requirement by M-Pesa continues to be a key oversight objective
of the Central Bank.
4) A number of critical issues and risks that have been reviewed include; - liquidity
management, settlement risks, the reliability of the system, the registration of users,
system audit trail, anti-money laundering measures and consumer protection issues
that could compromise the safety, efficiency, integrity and effectiveness of the M-Pesa
system. It is also instructive that the CBK has so far not received a report of any loss
or fraud through the system.
5) The Central Bank of Kenya has continued to oversee the service in line with these
recommendations and its Oversight Policy Framework document on payment
systems in Kenya which is available at the Bank’s website, www.centralbank.go.ke.
For instance, whereas the system transacted about Ksh 17 billion in August 2008,
the net deposit/residual value per customer (i.e. deposit less withdrawals) was Ksh
203 thus demonstrating that M-Pesa has not been regarded as an alternative bank
account with sums of money staying in the system.
6) To further provide a sound legal basis for payment systems in Kenya, the CBK and
the Treasury have been engaged in several legal and regulatory measures aimed
at promoting safety, efficiency and effectiveness of payment systems in Kenya.
One such effort is the review of the Central Bank Act in the year 2003 to include
section 4A1(D) that mandates the CBK to promote such policies as to best promote
the establishment, regulation and supervision of efficient and effective payment,
clearing and settlement systems. Currently, the Bank has proposed and formulated
the enactment of the National Payment System Bill that will strengthen the above
mandate by inter alia expressly providing for the oversight of all Payment systems
including money transfer services. This Bill will soon be tabled in Parliament for
enactment into Law.
7) It is also noteworthy that the recently enacted Kenya Communications (Amendment)

Act 2008 expanded the functions of the CCK in relation to electronic transactions
and provides legal recognition of electronic transactions. The Act not only legalizes
electronic transactions but it also enables the CBK and CCK to work together and
31
support this system including other such products that may come in future to the
market.
This audit by the Central Bank on M-Pesa system provides comfort to the Ministry of
Finance and I would like to assure Kenyans that this innovative idea of money transfer
through the mobile telephones is safe and the Treasury and the Central Bank will
continue to oversee its safety as the innovations in the system and outreach progresses.
MINISTER FOR FINANCE January 6, 2009
In a sense, the audit was the final potential handbrake on the growth of M-Pesa. Once
the service had jumped that hurdle, it could grow. And it did, contributing to five
generations. I turn to them now.

32
5
generations
The mobile money growth, supported by policy innovation from the CBK, has triggered
a wider financial services revolution in Kenya. I describe the progression and evolution
of the mobile money in Kenya under five generations here.
5.1 Generation 1: Initial mobile money growth and

financial inclusion
The starting point is to recognize financial inclusion as a public policy to fight
poverty. It improves the well-being of all the participants and makes markets
accessible. Increased access to financial services to the poor provides a safe haven
for their savings as long as those savings are safe and are not consumed by cost of
maintaining and servicing the accounts. Most of the poor are target savers and they
are efficient in saving-investment cycles. This allows them to widen their economic
opportunities, increase their asset base through increased savings and affordable
credit and therefore reduce their vulnerability to external shocks; this is savings for
consumption smoothing and savings to accumulate capital and thus sustainably
escape the poverty trap.
When M-Pesa was launched in 2007, financial inclusion, describing the percentage
of the populations’ access to formal financial services, was far lower than for a
developed country. Just 26.7% of the population had access to formal financial
services. Some unbanked and low-income communities had opened micro savings
accounts, but a trip to the bank either to withdraw or deposit was a very expensive
one. Therefore, the initial target was not to make them banked, but to ensure they
participated in the payments system of this new framework. M-Pesa allowed the first
entry into the financial system and later allowed the comfort of banking without
a trip to the bank.
M-Pesa has had a powerful effect on mobile money. Now, over 86% of the population
is within 5 kilometres of a financial access touch point (FinAccess Geospatial Mapping,
2015). Mobile money has underpinned this growth. There are now 30.5 million active
registered mobile money subscriptions in Kenya of which M-Pesa subscribers are
30.2 million while Airtel money and T-Kash are 0.344 million as at June 2020 (CAK Q4
report, 2020).
33
Other mobile money services emerged particularly after the 2008 audit discussed
in section 3. Now, the big banks and other MNOs realized they had been left out
by the market developments, including innovations that had boosted the market.
Other MNOs rushed to register and roll out their products: Zain launched Zain-Zap in
February 2009, Essar Yu mobile launched Yu-Cash in December 2009, Orange launched
Orange money in December 2010. Therefore, the mobile phone financial services
ecosystem was enlarged, and by the end of 2010 there were six (6) mobile money
providers and two (2) third party providers – Tangaza Pesa and Mobikash (see Table
5.1). Similarly, other commercial banks that had not sought to partner with Safaricom
to establish a similar platform like CBA (there were no restrictions on such moves)
after January 2009 started looking for avenues to participate in this ecosystem. It
was perhaps the realization that this revolution was unstoppable. The action by the
Acting Minister for Finance indirectly and unconsciously opened a new chapter for
DFS revolution, confidence and dynamism in the Kenyan financial market.
Table 5.1: The entry of the MFS providers in the market
Mobile Network Operator cum MFSP Product Entry Date

1 Safaricom (K) Ltd M-Pesa March 2007 - Current
2 Essar Telecom – Yu Mobile Yu Cash December 2009 – January 2015
3 Orange (K) Ltd Iko Pesa November 2010 - Current
4 Bharti Airtel Airtel Money September 2011 - Current
Mobile Financial Service Provider Product Entry date
1 Mobile Pay Ltd Tangaza January 2011 - Current
2 Mobikash Africa Ltd Mobikash July 2011 - Current
After the initial success of the mobile phone financial services, the risk mitigation and
the adoption/confidence in the market, the policy questions had to be addressed. First, I
argued that M-Pesa would form a technological platform for a menu of financial services
to be rolled out. That is the way I predicted from the available information. But then,
what were the initial goals of CBK and the Government in encouraging this product to
be developed and tested in the market? Have these goals changed over time or have
they been enhanced? To be mild on these issues, it does appear that financial inclusion
policy thinking as it was emerging, starting from microfinance whose legal framework
was six years behind schedule, and the bank branch network in the country were low
and not expanding fast, with the technology for managing micro accounts only being
implemented by few banks that were also dismantling the barriers to entry. All these were
given a new lifeline by M-Pesa platform. It would later shape further developments and
directions on financial inclusion in Kenya. It was a new lease of life for banks because it
provided them with an efficient way to effect payments. This then would bring customers
to the banks for transactions. It thus formed an efficient technological platform for
payments and managing micro accounts for smaller banks.
34
For the CBK, the goal was and still is to make financial markets accessible and the
commercial banks were the easiest entry point to the financial sector, but then what
was the problem? I have listed above the barriers to entry in the banking system.
M-Pesa was now a product that would overcome not only the barriers to entry erected
by commercial banks, but also solve the constraints of geography, physical distances
to the banks and provide a technology to manage micro accounts. Second, there was
need to fight informal financial services and the totality of informal markets because
M-Pesa as a product and as a payments platform was communicating across all
markets. We recognize now not only the benefits of financial inclusion but also the
benefits of formality of markets for regulators and national security issues, including
the AML/CFT regimes. Kenya is a typical African economy with segmented markets and
a preference for informal markets. The segmentation is driven by several characteristics
but income levels, irregular flows of income, cost of ‘formality’ - structural problems
and distance to the market are important factors that impose constraints on market
growth in size and location. Third, for the CBK, the only way to increase the capacity
for banks to grow in future was to encourage them to grow and increase their deposit
base at the time. There were only 4.7 million deposit/savings accounts and about
700,000 loans and advances accounts for an economy with over 30 million adults and
43 banks, as at December 2007. Only one or two banks had developed a technology
to manage micro accounts; this would take a long time to influence the totality of
Kenyans given the cost to roll out brick and mortar branch outlets. The commitment
by both the CBK and the Government to reduce the cost of doing business was then
to use the available channels to create a mass market that was efficient and to ensure
financial inclusion. That is why innovations in this area received wide approval by the
CBK even though other stakeholders were either unsure of the directions that were
shaping up or were protecting the available avenues that existed.
To summarize the policy direction at the time, financial development as the CBK
perceived it, was shaped by four factors:
1. Banks that would serve Kenyans to help them save and invest – these banks
required a large base of depositors to give them adequate capacity in the
intermediation process and dynamism of banking products to grow.
2. The strategy then was to encourage banks to grow their branch outlets in rural
areas and poorer urban locations and also at the lower end create a category of
microfinance institutions at the nationwide and at the community level.
3. The monetary policy regime depended on monetary policy instruments that

worked through banks to affect liquidity in the banking sector and its pricing, but
a large proportion of currency outside the banking sector made these monetary
instruments weak. Published papers on inflation (for example Ndung’u, 1994) had
shown that inflation was driven more in the short run by food, energy and growth
of outside money.
35
4. The Government’s Kenya Vision 2030 prescribed increased savings and

investments rates and that Kenya would become the regional financial hub.
Therefore, the starting point was regional diversity, deepen the financial market
and then target financial development goals. The process was thus not entirely
clear and required time, but it does appear that it was dislodged and given a quick
direction by the M-Pesa technological platform that worked from downstream to
bring the unbanked to the banks and upstream by allowing banks to grow their
capacity through deposits and savings accounts and increasing the activities in
the intermediation process.
Box 5.1: Pauline Vaughan, Wolfgang Fengler, Michael Joseph (2010) argue that
regulation followed innovation
An important factor in the success of mobile money in Kenya was the progressive role
of Kenya’s regulators, especially the Central Bank. Mobile money entered a regulatory
vacuum. At the time M-Pesa was piloted, no regulations existed for e-money type
initiatives, or for the involvement of mobile phone operators in any kind of financial
transactions. The operator kept the Central Bank updated on the developments, inviting
critique and suggestions through the pilot.
In preparation for a commercial launch of mobile money, Safaricom sought approval

from the Central Bank. Safaricom and the Central Bank worked together to address key
aspects of payment system regulation including product functionality, legal compliance,
stability and redundancy of the technical platform, prudential controls and consumer
protection. The Central Bank consulted with relevant governmental and policy bodies,
including DfID through their local representative Financial Sector Deepening (FSD).
In parallel, Safaricom lobbied the government and notably gained the support of the
Permanent Secretary in the Ministry of Information and Telecommunications
36
Box 5.2: Reflections by Michael Joseph on M-Pesa’s critical moments
I firmly believe that the decision to allow Safaricom to go ahead with M-Pesa was firstly
because no one really realized or anticipated the phenomenal success that M-Pesa
would achieve and the scale of the adoption, and secondly that all parties, particularly
the Central Bank, wanted to support innovation and agreed that regulation should
follow innovation.
The real key to the success of the M-Pesa rollout and acceptance was the number and
geographic spread of the M-Pesa Agents. Not many people, either within or external
to Safaricom, understood the concept and the necessity of the number and spread of
M-Pesa Agents. Thus, the necessity of getting regulatory approval for the management
and appointment of Agents was not sought as rigorously as for the product itself. It was
only after the initial success and the concerns that were raised by the traditional banks
that attention was then given to getting regulatory “approval” of the Agent structure.
The joint launch of M-Pesa by both Ministries of Finance and Telecommunications, the
support by the Central Bank Governor and the genuine innovative culture of Kenyans,
both within the Central Bank and Safaricom staff, were the basis of the huge success of
M-Pesa and the subsequent defense of M-Pesa by Treasury when the traditional banks
belatedly realized the potential impact on their own business.
Personally, I was determined to learn from some of the lessons of the past where I had
launched a new product but had not given it sufficient resources or attention to make it
succeed. Launching new value added services of such magnitude requires dedication,
passion, commitment and imagination even if you are not expecting (as we were) success
on the scale we achieved and even if the “business plan” tells you that you are crazy!
Source: Adapted from the Author
The starting point of M-Pesa provides an important background to the above four
factors and reflections that seem to document how regulations followed innovation
and somewhat complemented the gap filled by M-Pesa.
One of the points and perhaps questions emerging from the literature and observations
of the DFS in Kenya is that it has not formalized the seemingly large informal market.
This was not the objective for M-Pesa. It is important to note that the DFS is being used
by formal and complex market structures and informal markets. One can pay for a
meal in a five-star hotel using M-Pesa and also pay for a cup of tea in a roadside kiosk
using M-Pesa as well. A product such as M-Pesa being used across market segments
in Kenya lowers transactions costs considerably and can be considered a success,
but it is a matter of time before it becomes the main coordinating instrument across
markets both formal and informal. In the long run, the M-Pesa platform will collapse
the formal/informal divide.
37
We do not claim direct attribution but we argue that M-Pesa created the appropriate
environment for financial inclusion to take root.
The success of M-Pesa can be looked at in three dimensions. First, M-Pesa through
transfers and payments of goods and services supported the development of the
national payment system that has become effective, efficient and transparent. Second,
the M-Pesa technological platform has become an efficient and effective way for
both the micro savers and depositors in managing their bank account transactions.
Finally, the M-Pesa technological platform has developed further to allow a platform
for short term microcredit that has been efficient and effective without a trip to the
bank. Other countries from Africa, Latin America and Asia supported by Alliance for
Financial Inclusion (AFI) and other partners visited Kenya to learn how M-Pesa worked
and what they needed to do to replicate the same type of DFS platform in their own
countries. It is also worth noting that Kenya was one of the founder members of AFI in
2009 and at the time it felt that it had a contribution to make in the developing world
to fast-track financial inclusion. I was the Chair of AFI Steering Committee in the first
four formative years. Since then, AFI has become a premier network coordinating
financial inclusion policies in the developing world of Asia, Africa, Latin America, the
Caribbean and the emerging markets. Kenya is seen as a “clearing house” for financial
inclusion policy solutions that have worked, have been tested and can be replicated
in other developing and emerging economies.
Figures 5.1 and 5.2, and Table 5.2 show the growth of M-Pesa in terms of customers
and accounts from the time of the launch to the current. The three sets of information
show a combination of M-Pesa Agents growth, the customers enrolling to the system
and the level of transactions.

Figure 5.1: Total customers since March 2007 (millions)
38
In Figures 5.1 and 5.2 for example, it shows that the number of M-Pesa accounts
increased to over 25 million by 2015. The level of transactions had reached Ksh 206.1
billion per month in February 2015 and over 70 million transactions. By August 2020,
the number of mobile money accounts had increased to over 60 million, whereas the
value and volume of transactions had increased to Ksh 473.5 billion and 163 million,
respectively. The computed daily transactions in February 2015 was Ksh 7.4 billion
(US$ 80.5 million) per day and in August 2020 it had risen to Ksh 15.3 billion (US$ 145.5
million) per day. However, the average value per transaction has remained relatively
low at about Ksh 2,935 (US$ 32) for both periods in Feburay 2015 and August 2020,
implying the payment services has largely remained for small value transactions, as
was purposed when M-Pesa was launched in 2007.
Figure 5.2: Value and volume of M-Pesa transations, March 2007-August 2020
The total number of Agents by the end of 2019 was 224,108 and the number of active
customer accounts was 58.36 million compared to 6,104 Agents and 5.08 million
customers in 2008 (Table 5.2).
Table 5.2: M-Pesa agents, accounts and transactions

Year 2008 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Active Agents 1,812 2,067 2,329 2,606 2,770 3,011 3,378 3,761 4,230 4,781 5,399 6,104
Mobile money Customers (Mn) 1.59 1.82 2.08 2.37 2.72 3.04 3.37 3.73 4.14 4.42 4.75 5.08
Volume of transactions KSh (Mn) 1.35 1.74 2.40 3.07 4.02 4.20 5.39 6.34 7.15 8.30 8.57 10.21
Value of transactions KSh (Bn) 4.06 5.22 6.75 8.39 10.90 10.92 14.02 16.76 19.27 21.60 21.70 26.99
Year 2019
Active Agents 201,336 212,252 226,957 230,220 224,825 222,484 222,087 222,479 224,959 223,176 222,211 224,108
Mobile money Customers (Mn) 40.30 50.04 50.36 52.05 52.20 46.80 53.89 54.78 55.70 56.29 58.04 58.36
Volume of transactions KSh (Mn) 154.24 144.49 161.38 155.80 153.26 149.73 152.98 151.83 151.22 156.11 153.06 154.99
Value of transactions KSh (Bn) 368.02 328.15 368.39 360.22 364.25 346.85 366.39 368.50 365.91 366.90 359.26 382.93
39
This figure shows how the M-Pesa Agents and customer base has supported its growth.
The entry of other mobile financial services (MFS) providers in 2009 has actually
enlarged the market base. However, Safaricom Agents still dominate with market
share of 87.5% (221,333 Agents) while other Agents form 12.5% (31,370 Agents) of
the market share as at August 2020.
Figure 5.3: Mobile financial services agents
The results emerging from M-Pesa’s growth have impacted financial inclusion. Thirteen
(13) years of data points tracking reveals some interesting results for financial inclusion
in Kenya:
• The proportion of the adult population included in formal financial services has
increased from 26.7% in 2006 to 82.9% in 2019. Those preferring the informal
financial services have declined from 32.1% in 2006 to 6.1% in 2019.
• The proportion of the adult population totally excluded from financial services has
declined from 41.3% in 2006 to 11.0% in 2019.
• The proportion of women excluded has declined from 42% in 2006 to 27% in
2013, and further to 11% in 2019, the same level as the men excluded as at
2019. The acceleration seems to coincide with M-Pesa and a whole range of
accessibility to the financial system between 2009 and 2019 where data points
are comparable.
• The men’s profile of financial inclusion has been relatively better. Those excluded
have declined from 41% in 2006 to 24% in 2013 and declined further to 11% in
2019.
40
Table 5.3: Financial inclusion profile in Kenya 2006-2019 (% of adult population)
Financial access Total % Rural % Urban % Male % Female %

category
2006
Formal 26.7 23.8 35.5 33 21
Informal 32.1 35.5 21.6 26 38
Excluded 41.3 40.7 42.8 41 42
2009
Formal 40.5 34.6 62.4 48 39
Informal 26.8 29.5 16.5 20 39
Excluded 32.7 35.9 21.1 32 39
2013
Formal 66.7 59.6 80.0 71 63
Informal 7.8 9.8 4.3 5 11
Excluded 25.3 30.6 15.8 24 27
2016
Formal 75.3 69.0 86.3 80 71
Informal 7.2 9.0 4.1 4 10
Excluded 17.4 22.0 9.5 16 19
2019
Formal 82.9 77.3 91.2 86 80
Informal 6.1 8.3 2.8 4 8
Excluded 11 14.4 6.1 11 11
Source: FinAcess Survey (2019)
• In terms of rural-urban divide, financial exclusion has followed the national

average, but urbanites have better financial access than their rural counterparts.
By 2019, only about 6.1% of the urban adult population was financially excluded
compared to 14.4% of the rural population.
• The preference of mobile phone financial services across rural/urban and across
gender and age cohorts seems to explain the financial inclusion and accessibility
of financial services.
The Kenya’s National FinAccess Survey 2019 revealed that 32.7% of Kenya’s bankable
population is totally excluded from both formal and informal financial services. If
46.0% are poor and 32.7% are unbanked, then most of the unbanked are the poor.
Strategies to enhance financial inclusion were therefore seen as critical in efforts to
reduce poverty and uplift the bottom billion on a sustainable basis. The FinAccess
Survey 2006 had even more depressing statistics. From the 2009 survey, we can thus
trace the impact of this policy process and the success that has been seen to date.
Figure 4.4 provides a summary of the profile of financial inclusion in Kenya.
41
Figure 5.4: Progress in financial access in Kenya, 2006-2019 (%)
The results from this figure can be analyzed from three dimensions. First, the adult
population excluded from financial services declined from 41.3% in 2006 to 11.0%
in 2019. Second, the importance of informal financial services significantly declined
from 32.1% of the adult population preferring informal financial services in 2006 to
6.1% in 2019. It shows that once formal financial services are accessible, preference
for informal financial services declines. Finally, looking at the proportion of the adult
population served by formal financial services, it has increased over time from 26.7%
in 2006 to 82.9% in 2019. These results show that from whichever dimension one looks
at, financial inclusion has been a success story in Kenya. In addition, financial access
touch points have been expanding. Kenya is ahead of its peers (Figure 5.5). There is
an increase in bank branches, Automated Teller Machines (ATMs), Telco Agents and
Agency network for banks that have increased to over 35,000 since inception in 2010.
Figure 5.5: Financial access touch points
Source: Country Geospatial Surveys, 2013
42
Figure 5.5 has three sets of information on accessibility of financial services. First
on the physical distance, second on the number of access points and finally on the
access points per given population size. The results show that Kenya is ahead of its
comparable peers on financial inclusion. Further, in the most recent FinAccess Survey
of 2019, Kenya ranks third to Sychelles and South Africa (Figure 5.6).
Figure 5.6: A regional comparison
Apart from Kenya, growth in usage of banking services is limited. The largest growth
driving formality in most countries is the usage of other formal services, which in most
cases include mobile financial services just as in Kenya and Agent banking. Of the
10% growth in banking services recorded between 2013 and 2016, 6% is accounted
for by new mobile banking products such as M-Shwari and KCB-Pesa. The growth in
formality in Rwanda is worth noting. This has more than doubled in just four years
from only 19% in 2012 to about 68% in 2016. The growth has largely been driven by
both digital financial services and Savings and Credit Cooperative Societies - SACCOs
(FinScope, 2016). This expansion is in turn attributed to the Government of Rwanda’s
commitment to implementation of national savings mobilization and digitization of
government services.
43
5.2 Generation 2: Linked Bank Deposits and

Partnerships with Financial
Institutions
The second generation saw more investments and new and innovative products
emerge8. Therefore, the next phase of development was virtual banking system,
starting with savings accounts operated under the same platform but now being
individualized savings accounts; that is, an M-Pesa account holder would have a direct
link to a personal savings account in the same M-Pesa menu in the phone. But more
imporatntly, account holders would be able to use the electronic retail payments to
complement financial services, with the emergence of ‘Lipa na M-Pesa’ (pay with
M-Pesa) in all retail and wholesale outlets of goods and services.
A key point here is the response and later engagement of banks and other financial
institutions with M-Pesa and other mobile money systems that emerged. To move
forward with some consistency, I look at the main stakeholders in this game:
• The multinational banks reaction to M-Pesa: The multinational banks looked

at M-Pesa as a major disruption. Most of these banks were serving and still serve
corporate clients and trade financing. But they viewed M-Pesa as an instrument
that would drive liquidity from all other banks to the bank holding the payments
platform, the CBA. They viewed it as purely loss of business in the interbank market
for domestic liquidity where they dominated on the selling side. They therefore
warned me of a massive banking crisis in Kenya due to liquidity exodus from all
other banks to one small bank holding the trust account, the payments solution
platform. They never looked at M-Pesa providing solutions to the payment system
and that they would be part of that payment system.
• The leading Kenyan commercial banks and microfinance-based banks:

The leading Kenyan banks and especially those that had just perfected the
technology of managing micro deposit and savings account led a revolution in
removing minimum balances and other restrictions on savings accounts. They
viewed M-Pesa as an alternative account, so the SIM card was in competition
for micro savings and deposit accounts. These banks looked at the convenience
8 I recall having breakfast one morning of September 2008 in Nairobi with Bill Gates and Michael
Joseph and how impressed Bill Gates was with M-Pesa and that it had the potential to change
peoples’ lives in Kenya and beyond. But then he argued that this success was not affecting the
banking intermediation process and also would make lasting impacts on peoples’ lives if they
could save in the banking system but not with the then high transactions costs of transfers from
M-Pesa account to a savings account. We assured him that this was the next phase of M-Pesa
development. As of now we recognize not just a phase but subsequent incremental and successful
phases of innovative developments that have changed peoples’ lives
44
M-Pesa provided, the transparency of costs and effectiveness of charges, hence

they envisaged a situation where there would be a massive exodus from micro
accounts holders to M-Pesa. My assurance to Kenyan banks and deposit-taking
microfinance institutions was that M-Pesa was an appropriate instrument or
product (depending on what they wanted to use it for) they should integrate
with. In doing so, their customers would not need to visit the bank to withdraw
or deposit; they would thus reduce tellers in their banking halls and earn ledger
fees 24 hours a day for seven days a week. In short, I advised that M-Pesa would
become a better technological platform to manage micro savings and deposit
accounts.
• The CBK, the Ministries of Communication, and Finance and subsequent

developments in M-Pesa: The Ministry of Communication was happy that the
telcos were taking part in the money transfer business and that working with
and advising the Communications Authority (CA) had made the difference. But
in the Ministry of Finance, for the majority of the senior staff it was purely a CBK
business and if anything went wrong, the CBK would be blamed, not the Ministry
of Finance, save for the support by the Minister, Amos Kimunya, who launched
M-Pesa. As shown so far, in the end there were winners only and no losers. Section
2, outlining the internal audit, shows how this audit was addressed.
By using banking data and by tracing deposit and savings accounts growth and banks’
branch outlet growth, we show that the market expanded in both branch outlets
and deposit accounts. These developments are also supplemented by the growth of
microfinance banks and agency banking.
5.2.1 M-Shwari
We provide micro-evidence, due to data challenges, by tracing one bank, the CBA
from its M-Pesa platform that revolutionized the payments system in Kenya, to a
virtual savings accounts platform, M-Shwari, allows savers and records savings
and transactions data to form a set of credit scores for pricing credit for each
customer. This bank has built virtual savings accounts in 40 months since M-Shwari
was launched to cover over 14 million customers, a substantial savings network
and forms a base for rising credit demand for short term credit. By January 2019,
the customer base was 28.8 million. This micro evidence is also available in CBA’s
branch outlets in Tanzania (M-Pawa product), a virtual savings and credit supply
product like M-Shwari. This provides further evidence of regional impact of DFS in
the Eastern African region.
The M-Shwari product was launched by CBA in November 2012 to provide virtual
banking services to mobile-centric customers for both savings and loan products.
45
The objective was to provide simple and affordable financial services through easily
accessible channels. It was a visionary product of CBA in partnership with Safaricom,
and just like M-Pesa it has made a major mark. The females using the savings product
is higher than the national average and the youth (34 years and below) dominate the
savings product and credit demand.

Table 5.4: Creating a virtual savings base: M-Shwari
Year Customers Deposits (Savings) Micro Loans

2012 4,786,762
2013 5,026,955
2014 7,000,113
2015 12,459,833
2016 (Jan.) 17,293,055 US$ 80,935,428 US$ 914,499,746
2017 21,462,163
2018 25,842,902
2019 (Jan.) 28,782,898
Source: Commercial Bank of Africa, 2019
After 40 months of operation, in January 2016, total deposits stood at US$ 80,935,428.
The average savings was as low as US$ 5.3 per month. The total loans disbursed stood
at US$ 914.5 million. The average number of loans disbursed stood at 67,937 and
the average loan size at US$ 31 and the average loan repayment period was 26 days.
The structure and characteristics of savers in M-Shwari is also quite unique. Of the
14.025 million customers, men are 7.35 million, which is 59% of the total customers
and women are 5.12 million and which is 41% of the total customers in this product.
With regard to age cohorts, those in the age of 18-24 years comprise of 29% of the
total customers, those in the 25-34 years age category comprise of 38% and those
in the 34-55 years bracket comprise of 25% and those over 55 years are only 8% of
the total M-Shwari customers. This shows that 67% of those participating in this
mobile-centric banking solutions and taking loans are young people below the
age of 34 years. From the time of the launch, this product has registered 10,994
customers per day.
In addition, most M-Shwari customers are target savers and they have a facility to lock
their savings for a defined period. The customers with locked savings as at January
2016 stood at 147,213, with an average US$ 70 of locked savings for an average period
of 3.8 months. The non-performing loans proportion for this facility stands at 1.92%
compared with the national average of 5.3%.
One interesting outcome of this micro impact of M-Pesa and its progression to the
five stages of DFS revolution is that it is embodied by this bank, the CBA. It shows
46
that its partnership with Safaricom and their investment in the payments platform
in the initial stages has paid off. The subsequent developments have pulled the
totality of the financial evolution with dynamism and efficiency not only in Kenya but
also in Tanzania. In Tanzania, the equivalent of M-Shwari is M-Pawa (consistent with
empowering the poor to save in a bank account and to acquire short term credit).
Finally, the total micro-accounts (those with Ksh less than 100,000) in the banking
sector stood at 41.67 million in 2016, and CBA held 17.2 million of those accounts.
The CBA held slightly over 40% of the total micro-accounts in the banking sector in
Kenya in 2016. This, as we have seen, are massive deposits from target savers that
provide the bank with capacity to grow in future and participate strongly in the
intermediation process.
M-Shwari and similar products have moved much further, triggering wider bank
expansion. The debate on financial inclusion, product design and method of
financial services delivery and financial development were now gaining ground not
only in Kenya but in other countries with successful DFS evolution, such as Tanzania.
In Kenya, the initial fears and regulatory gaps, it appears, had been appreciated by
Kenyans. It was no longer an activity outside the mandate of the Central Bank of
Kenya.
The success of virtual savings accounts did not also deter the growth of branch outlets;
the emerging outcome seems to have been encouraging both physical branches and
virtual accounts. We show this in Figure 5.7 on the number of bank branches and the
distribution across rural and urban areas in Kenya. The branch outlets have increased
from 534 in 2005 to 1,490 in 2019. The rural branch network has not been left behind,
either growing from 181 branches in 2005 to 660 in 2015 while the urban branches
have grown from 353 to 783 over the same period. The acceleration of branch outlets
seems to start in 2007-2008 period. From 2015, official reporting of branch networks
by commercial banks changed from ‘rural and urban’ to reporting branch networks
by ‘counties’.
Figure 5.7: Growth in branch network in Kenya
47
Figure 5.8 shows the growth of deposit accounts in the same period, 2005-2019. Deposit
accounts have increased from 2.55 million in 2005 to 64.7 million in 2019 with over 95%
of these deposit accounts being micro accounts, and currently 94% are micro savings
accounts. The number of micro accounts has increased more than twenty five-fold
from about 2.1 million accounts in 2005 to 63.1 million accounts in 2019. This growth
is attributable to reduced costs of maintaining micro accounts and introduction of
innovative instruments targeting lower tier market segments – using the M-Pesa platform
to open savings accounts such as M-Shwari, KCB-Pesa and other DFS products.
Looking closely at the growth of micro-accounts, the acceleration seems to start

in 2007 from 4.12 million accounts to double to 8 million accounts in 2009. The
progression from then on is driven by the dynamics in the market; it is also the
period that M-Pesa was registering 10,000 accounts every day. By 2010, one of
the complementary additions to the branch network was the successful launch of
Agency banking. Two banks started to appoint Agents in 2010 and so far 16 banks
have appointed over 3,500 Agents (Figure 5.9). The pattern of branch networks, the
success of the agency model and the success of M-Pesa all seem to corroborate the
pattern of deposit accounts and the vibrancy of the banking sector in Kenya over
this period.
Figure 5.8: Growth in bank deposit accounts in Kenya, 2005-2019
The number of institutions with bank Agent networks rose from 2 in 2010 to 16 in
2014 and the number of commercial bank Agents has increased to more than 35,000
since the launch of the agency banking initiative in May 2010. As at 2019, the number
of Agents had increased to 67,314. In addition, the value of transactions handled by
Agents had risen from US$ 469 million in 2011 to more than US$ 3.7 billion in 2014
and close to US$ 12 billion in 2019. This growth is attributable to penetration of bank
Agents in underserved areas, and the increase in financial access touch points.

48
Figure 5.9: Agency banking uptake in Kenya, 2010-2019
Other avenues and reform measures by the Central Bank have also complimented
the DFS to raise the financial inclusion profile for Kenya in the period. They include:
• First, licensing of Deposit-Taking Microfinance (DTM) institutions both nationwide

and community micro finance institutions that later changed their name to
Micro-Finance Banks (MFBs). They provide financial services at a closer reach to
low income segments. The outreach introduction of Agency Banking model later
helped the MFBs by easing branch requirements/specifications.
• Second, the commercial banks were encouraged to expand the branch network/
outlets. This increased commercial banks’ branches from 534 in 2005 to 1,113 by
end of August 2011 to 1,490 in 2019. Growth has been driven mainly by competition
and declining barriers to entry.
• Third, Agency Banking – Turning non-bank outlets into financial services providers.
So far, 67,314 approved bank Agents facilitating 3.9 million transactions valued
at Ksh 1.2 trillion (US$ 12 billion), leveraging on mobile phone Agents also. These
Agents have pushed forward financial inclusion frontiers as main access touch
points for financial services and serve several banks (as there are no exclusivity
clauses).
• Fourth, increasing the core capital of banks to enhance strong banks that raises
the level of confidence in the market and becomes a safe haven for savings. The
enhancement of core capital for commercial banks to a new level of Ksh 1 billion
(US$ 10 million) from the level of Ksh 250 million (US$ 2.5 million).
• Fifth, reducing cost of doing business - Currency Centres - Reducing cash in transit
costs for banks and their branch networks across regions.
49
• Finally, consumer protection - the missing link in financial inclusion. Consumer

protection rights were entrenched in Kenya’s new Constitution. The new
Constitution gave a legal framework to develop strong institutions to grow the
market.
The effects of M-Pesa, linked bank accounts (to M-Pesa) and agency banking go further,
triggering wider financial inclusion in Kenya. In the 2019 FinAccess survey, 79.4% of
the adult population used the mobile phone-based financial services compared to
61.6% in 2013. The highest growth was in the use of digital Apps loans that grew by
8.3% in 2019. The uptake of mobile banking increased from 17.5% in 2016 to 25.3%
in 2019 while traditional banking declined from 31.7% to 29.6% over the same period
(Figure 4.10).
Figure 5.10: Utilization of financial services across the providers: 2006-2019
Barriers to entry into the financial system have been significantly reduced, leading
to increased usage of financial services across all providers (Figure 5.11). The rise is
most notable in the use of banking and insurance services. The proportion of adults
with access and use banking services almost tripled from 3.9 million to 10.2 million
while those accessing insurance quadrupled from 600,000 to 7 million during the last
decade (2009-2019). The number of adults using mobile money increased to 19.9 million
in 2019 from 5.3 million in 2009.

50
Figure 5.11: Adults using financial services providers (million)
5.3 Generation 3: Linked credit scores
The third phase that quickly became a natural progression was short term credit
applied and disbursed through the same platform. The novelty of this phase was
that banks and telcos invested in a more versatile platform that used the transactions
and savings data to generate individual credit scores and customize the price of
credit, thus emerging as the central platform for digital credit assessment and the
distribution of micro-credit (Ndung’u, 2018). This in turn would revolutionize the
collateral technology that has acted as a major barrier to borrowing from the financial
system by many potential investors; that is the credit market has been hampered by
the existing costly collateral technology.
Another key development to use of mobile phone financial services was access of
micro-credit anytime anywhere within network reach. At 3:00 a.m one could borrow,
trade in micro business (selling vegetables in Wakulima market) and by 7:00 a.m,
repaid the loan and earned a profit. Access to credit was no longer confined to a
building structure-bank that was time consuming with no immediate results. This
access of credit by small savers and borrowes was key in enhancing operations of
small businesses in Kenya, which contribute 28.4%9 to the gross domestic product.
Technology thus made it possible to access credit anywhere and anytime, which
enhanced economic activity.
9 Kenya National Bureau of Statistics, Kenya-Small and Medium Enterprises (MSME) Survey 2016
51
The development led to commercial banks partnering with other mobile money
transfer and mobile money commerce operators to offer credit. Besides the pioneer
Safaricom’s M-Pesa platform, we have KCB-M-Pesa, M-Coop by Cooperative bank,
Timiza by Absa bank, M-Shwari by CBA, Airtel (Airtel Money), Telkom (T-Kash), and
Mobile Pay Ltd (Tangaza). Equity bank also developed its own M-money system –
Equitel. Moreover, increase in demand for credit and high preferences for technology
(mobile phones) has led to proliferation of other digital lenders in the Kenyan
financial market. As at December 2018, there were 23 digital credit providers that
were unregulated by the Central Bank.
Two pieces of evidence from FinAccess 2019 and the CBK reinforce on the point
of preference to use MFS rather than banks as a source of credit. Figures 5.12 and
5.13 show an uptake of digital accounts ownership and registration that increased
significantly in 2019 compared to 2016, reflecting high adoption of digital accounts.
Those with access to a mobile phone, own or borrowed, increased to 91% in 2019
compared to 78% in 2016.
Figure 5.12: Active digital accounts in Kenya 2016 versus 2019
Figure 5.13 shows that compared to Uganda and Tanzania, Kenya is ahead on digital
account usage for savings, borrowing and transacitons purposes.
52
Figure 5.13: Active digital accounts, country comparison
The lending to Micro Small and Medium Enterprises as a percentage of total banking
industry loans has declined over the years from 23.4% in 2013 to 15.8% in 2018,
implying the preference of MSMEs to borrow digitally from other credit providers as
opposed to borrowing from the banks. However, in value terms, the amounts have
increased to Ksh 393.0 billion in 2018 from Ksh 332.0 billion in 2013.
Table 5.5: MSMEs lending compared to total banking sector loan portfolio (Ksh
billion)
As at December MSMEs Loan Total Banking Sector MSMEs Loans/Total
Portfolio, KSh.Bn Loan Portfolio, KSh. Loan Portfolio%
Bn
2018 393 2,487.34 15.8
2017 413.9 2,155.73 19.2
2013 332 1,418.80 23.4
2011 225 1,076.56 20.9
2009 133 682.05 19.5
5.4 Generation 4: International remittances
The fourth challenge in the expansion of the M-Pesa technological platform was to
enable cross-border payments and international remittances. The immediate result
of this development would be the transformation of Kenya’s informal Hawala money
transfer system to a formal money remittance system. Demand for regulations to
cope with these innovations and the more intensive use of technology to monitor the
money-transfer market have improved the AML/CFT regime. However, the expansion
of commercial bank branch networks, mobile network operations and mobile money
53
platforms across other countries in Africa and the world to enable cross-border
and international remittances pose an increasing challenge for regulators. Mwega
(2014) points out that financial integration implies that the negative externality costs
associated with bank failure go beyond national borders, a reality that is not taken
into account by national regulators and supervisors. Furthermore, the emergence of
virtual currencies, such as Bitcoin, pose a challenge to the central banks charged with
monitoring and supervising international transactions whose demand and supply
mechanisms are not yet clear.
Kenya’s cross border remittance transactions have been revolutionized through the
use of technology in mobile money transfer products. The adoption of technology
through use of M-Pesa has enhanced the flow of international remittances at lower
prices, facilitating money transfer to millions of low income and unbanked populations
in the rural areas in Kenya. Figure 4.14 shows that in Kenya, remittances have steadily
increased at an average annual rate of 14.3% in the last decade, rising from US$
570 million in 2006 to US$ 2.8 billion in 2019, constituting 2.9% of GDP. The steady
increase from 2008 to 2019 is attributed to adoption of technology (mobile money)
and plaforms–World Remit and TransferWise (by Equity Bank) that improved efficiency
and lowered the cost of transfer of money across borders.
Figure 5.14: Migrant remittance inflows (US$ millions)
Source: World Bank Migration and Remittances data, Central Bank of Kenya
Evidently, the presence of mobile money platforms has increased remittances inflows
in Kenya. The growth and levels of remittances inflows in Kenya has surpassed the
levels of foreign direct investment and portflolio equity flows. Yet, I am wary of the
fact that the statistics reflected in Figure 5.14 are the flows through formal channels,
54
which I believe grossly understate migrant remittances since most migrants use the
informal channels to send money, which is unrecorded. Kenya is one of the top eight
highest remittance-recipient countries in Africa after Nigeria, Egypt, Morocco, Tunisia,
Ghana, Algeria and Senegal.The international remittances flows have increased
financial depth and inclusion in Kenya.
However, there is need for development of regulation that would see reduction in cost
of sending money within African countries. Apparently, it is cheaper to send money
from countries out of Africa into Kenya. For example, the cost of sending US$ 200 from
South Africa and Tanzania to Kenya is 14.3% and 13.1%, respectively, while sending the
same amounts cost 6.6% from the United Kingdom. Huge variations also exist across
corridors as indicated by the cost of sending similar amounts of money from Rwanda
recorded at 5.5%, yet the neighbouring Tanzania sends the same amount of money at
double the cost (see Figure 5.15). Among the solutions to driving this cost down is use
of modern and cheaper technology to send money, adoption of cross cutting regulation
for countries in Africa, and also development of more payout stations in the rural areas
or where there is none.
Figure 5.15: Cost of sending money to Kenya by corridor
Source: Remittance Price Worldwide and Send Money Africa
5.5 Generation 5: Partnerships with Fintech Firms
The uptake of FinTechs in the financial industry has been enhanced by increased
uptake in the use of mobile networks and mobile embedded systems, the use of big
data and data analytics, and cloud computing, and crowd funding (Gai et al., 2018).
FinTechs may be concentrated on some aspects of the industry such as the mobile
sites for internet banking, or replace an established mode of delivery of services,
for example, independent online banks, mobile financial services like in the case
55
of Kenya. The fintech activities have permeated the whole of the financial industry
that includes loan technology, payments, international transfers, personal finance
and asset management, blockchain, the capital markets, equity crowdfunding and
security technology.
One question to ask is whether M-Pesa is a fintech product? I argue it is. FinTechs
perform three essential roles in the financial industry. These are: reduction of
transaction costs and enhancing efficiency; improve business and risk management;
and financial inclusion, which I have already demonstrated (see earlier chapters
including this one) using the M-Pesa Kenyan example.
• FinTechs reduce the cost of data collection and processing that makes it possible
to process huge volumes of transactions within seconds.
• FinTechs use available data/information to manage risks resulting in improvement

in businesses.
• The use of internet and mobile technology allows easier and cheaper access to
finance as a result of reduced costs of data collection and monitoring that lead
to improved efficiency that make access to finance more inclusive.
M-Pesa as a FinTech product has become a platform of financial services and products
used to achieve fruitful ends. It has worked as a retail electronic transactions platform
from DFS to digitization and has allowed Fintechs to roll out sustainable business
models. These business models cut across all the sectors of the economy. Some
examples in Kenya include:
(i) One Acre Fund – Raised productivity and incomes for smallholder farmers.
(ii) M-Kopa on domestic solar energy supply.
(iii) Water vending machines for urban slums for poor households.
(iv) Crop insurance models, virtual health insurance products (M-Tiba on financing
health services).
(v) JUMO, a non-bank credit-only lender, now partners with a range of mobile money
firms in Kenya
56
In Kenya, mobile money facilitates related innovations in Kenya’s payment system,

particularly the Government’s e-citizen initiative.10 The Kenya Government delivers
a wide range of services, including birth certificates, driver’s licenses and passports,
electronically and collects payments of fines and penalties. The Kenya Revenue
Authority has designed tax payments platforms based on retail electronic payments
platform. Card payments are accepted, but mobile money payments comprise
an estimated 99% of the volume of e-citizen transactions.11 Mobile money also
facilitates activities in Kenya’s capital market. For example, Safaricom has offered
shareholders the option of receiving dividend payments through M-Pesa. Since March
2019, the Kenyan Government has offered M-Akiba bonds. This is a retail offering
sold in minimum amounts of Ksh 3,000 (US$ 30) via the Safaricom and Airtel Mobile
Network Operators (MNOs) networks.12 It provides an interest rate of 10% with no
taxes, providing small services with returns far higher than bank savings products.13
Further, FinTechs have also laid their strong foundation in the market, which has
caused banks to compete aggressively for mobile banking customers. The rapid uptake
of mobile phones has also fuelled the uptake of FinTechs. A survey carried out on
banks worldwide in 2017 by Statista asking them whether FinTechs posed challenges
or opportunities for the sector showed that a majority believed that fintechs offered
more opportunities than challenges. Over 90% of the bank executives responded
that they (banks) were willing to collaborate with FinTechs; 4% said they will acquire
them and another 4% indicated that they will compete with them.
Figure 5.16: Future strategies of banks regarding fintech companies worldwide

2017
Source: Statista
10 Kenyan Government < https://www.ecitizen.go.ke/ecitizen-services.html>.

11 GSMA <https://www.gsma.com/mobilefordevelopment/programme/mobile-money/kenya-went-
digitising-person-government-p2g-payments/>.
12 Kenyan Government, Central Depository and Settlement Corporation, Nairobi Securities
Exchange <http://www.m-akiba.go.ke/M-Akiba-English-Prospectus.pdf>.
13 Kenyan Government, Central Depository and Settlement Corporation, Nairobi Securities Exchange
<http://www.m-akiba.go.ke/M-Akiba-English-Prospectus.pdf>.
57
5.6 Generation 6: M-Pesa impact on monetary policy

and beyond
The benefit of the financial innovation go further than financial inclusion, leading
even to a better environment for monetary policy. Figure 5.17 shows declining
velocity starting in 2007 but a significant decline from 2009. Figure 5.18 supplements
this information by showing the behaviour of money multiplier in this period. The
declining velocity and the unstable money multiplier would imply that the money
demand is unstable in this period, and this had implications for the monetary policy
framework.
Figure 5.17: Velocity of money
The evolution of the velocity of money depicts different regimes – financial

development and less cash changing hands. In addition, the rising money multiplier
may imply that the CBK may have lost control of money supply process, but actually
reflects financial innovation in the market. Consequently, targeting broad money via
reserve money as an intermediate target was inadequate and perhaps obsolete. The
monetary policy framework had to change in 2009 and CBK adopted Net Domestic
Assets in place of Reserve Money. In the more recent past, monetary policy framework
has been price-based.
58
Figure 5.18: Money Multiplier (M3/RM)
In addition, we may conjecture that the declining velocity of circulation of money

and significant decline of the proportion of currency outside banks may be reflecting
financial deepening and increased financial innovation from 2009 to date. The
declining currency outside banks and the significant velocity decline reflects changes
in behaviour of holding cash; people are keeping less and less money outside of
banks and preferring less cash in their daily transctions (Figure 5.19). The traditional
monetary policy interventions impact through well established and stable channels
where it is assumed that velocity and money multiplier were stable and predictable
in Kenya was thus violated.

Figure 5.19: Decline in cash outside bank in relation to broad and reserve money
59
Kenya’s financial depth has been rising with increasing financial inclusion as
shown by a rise in the ratio of broad money (M3) to gross domestic product in Figure
5.20. It rose to 37% in March 2020 from 35% in March 2007, peaking at 44% in June
2015. The growth of private sector credit (PSC) has followed a similar pattern as the
growth in broad money, increasing from 22.3% in March 2007 to 26.6% in March 2020,
and peaking at 35% in September 2015.
Figure 5.20: Increase in financial depth in Kenya (%)
To position these results on financial inclusion and supporting policy reforms that
have supported market developments, we have provided a background of Kenya’s
banking sector in the period 2007 to 2019. We have shown that Kenya’s financial sector
has undergone significant transformation driven by the DFS evolution. The DFS has
coordinated all market segments and in the last decade or so, we have witnessed:
• Significant decline of barriers to entry to the financial sector via removal of

minimum balances.
• Significant decline in cost of maintaining micro accounts - via ledger fees, etc.
• The introduction of new instruments and financial products targeting lower

segments of the population.
However, physical distances remain an obstacle; that is, where digital finance is
required. A trip to the bank is very expensive and most trips are not necessary to
deposit or withdraw but to perform some transactions or even to enquire on the
balances of their accounts. The poor in particular are transactions heavy, therefore
the uptake and use of mobile financial services was the best solution.
The financial inclusion picture that has emerged is consistent with the fact that DFS
revolution has allowed accelerating financial inclusion in Kenya where new financial
60
services and products have emerged, encouraged by innovation and sound regulation.
New delivery channels have been developed and these channels are effective,
transparent and efficient, and that payments system has emerged that has reached
the poor and the wealthy in the same speed. These services and products transcend
all market segments and, as shown from the FinAccess data, the use of mobile phone-
based financial services accelerated from 27.9% in 2009 to 61.6% in 2013 and from
71.4% in 2016 to 79.4% of the adult population in 2019. Other financial services seem
to have been complemented by the mobile financial services:
• The proportion of the adult population using banks has increased to 40.8% in 2019
compared to 29.2% in 2013 and 14.0% in 2006 before the mobile phone financial
services were introduced in Kenya.
• Insurance is serving 27.9% of the adult population using the 2019 survey compared
to 4.9% in 2006.
• The savings products in the banking sector (M-Shwari and KCB-Pesa) have
attracted 18% of the adult population – this is virtual banking service. The survey
results show that more Kenyans are now using mobile financial services and
mobile banking on a daily basis.
• Bank deposit accounts have increased from slightly over 4 million accounts in
2007 to over 55 million accounts in 2018.
• Financial access touch points have been expanding. Kenya is ahead of its peers in
financial inclusion as of the 2014 survey: (i) 76.7% of the population are within 5km
of financial access points compared with 35.1% in Tanzania, 42.7% in Uganda and
47.3% in Nigeria; (ii) Financial access points per 100,000 people stands at 161.9 in
Kenya compared with 48.9 in Tanzania, 63.1 in Uganda and 11.4 in Nigeria.
The conclusion from these outcomes is that M-Pesa has enabled a turnaround in
Kenya on financial inclusion; it has also been able to catalyse the entire payments
system development and has allowed communication/coordination across market
segments and informal markets.
It may be appropriate now to scheme with clarity and to produce some data points
to show the impact after more than a decade of M-Pesa. The following conclusions
can be supported by data about M-Pesa:
1. A financial inclusion platform has been developed and is likely to be more dynamic
in future but it has also created an endogenous demand for a national strategy on
financial inclusion to be developed, and that strategy has cut across the totality
of the financial sector.
61
2. There have been significant reductions in transactions costs in the economy; a

new revolution in the payments and settlement platform that is efficient, effective
and in real time has been developed.
3. The M-Pesa revolution and hence the DFS has created an internal demand for
completing the financial infrastructure. This involves developing institutions that
support and protect the financial market. Such institutions have included those
that provide information, such as credit reference bureaus that will in future use
this information capital to change the collateral technology in the country. In
addition, this has allowed space for deposit insurance mechanisms and finally
the Financial Reporting Centre to improve the AML/CFT regime. In addition, new
legislation that includes competition policy, consumer protection and financial
literacy have been considered.
4. The Kenyan commercial banks (not necessarily big banks) have used the DFS
as a cost effective platform to manage micro savings/deposits accounts and
microcredit supply and also to roll out products consistent with their market
niches. This DFS platform has enabled commercial banks to manage micro
accounts and, therefore, huge amounts of deposits from micro savers has given
these banks the capacity for intermediation in the market and to expand their
reach.
5. An efficient AML/CFT regime that is easy to monitor has reduced the role of informal
markets and recognized that financial exclusion is bad for AML/CFT regime.
6. An environment for monetary policy to work and a landscape to change monetary

policy framework now exists. In Kenya, as soon as M-Pesa success started to
be noticed in 2008, monetary aggregate targeting framework was considered
obsolete, the demand for money shifted, innovations in the financial market
increased the money multiplier, and the changing preference for holding cash to
cashless induced a drastic decline of velocity of money.
7. The political economy cannot be forgotten; the DFS has generated massive
economic rents and being distributed through this platform – at present about
4.5% of Kenya’s annualised GDP is transacted daily – that is Ksh 7.5 billion or US$
75 million per day.

62
6
lessons, challenges and
the future for digitization
We have shown how frontier developments and endogenous innovations shaped the
Kenyan market through a unique “test and learn” process adopted by regulators. What
started as a bank and telco product (M-Pesa) on a simple money transfer technological
platform has pushed innovation and the regulatory space towards important
digitization developments. Digital development in Kenya has contributed towards
strengththing state capacity in important ways. At the same time, these impressive
achievements have also raised concerns about the need to further strengthen state
capacity to cope with regulatory challenges, provide market safeguards, and evolve
a national identification system that is suitable for the next phase of an increasingly
digital society. This section highlights the main lessons learnt from Kenya’s experience
through the use of M-Pesa and the challenges facing the future digital environment.
6.1 Lessons drawn from Kenya
First, digitization has pushed the retail electronic payments system to cover virtually
the whole economy, including government services, pushing the government to
embrace an electronic payments ecosystem. One spillover effect has been the
formalization of most informal transactions, raising hopes that informal markets in
Kenya, and indeed Africa, will one day shift towards formality. M-Pesa-type or mobile
money or digital-based products have led to a vibrancy of financial markets cutting
across all sectors and improved transactions at all levels from the formal to informal
markets in Kenya and the East African region.
Second, Kenya has demonstrated that digitization can enable financial inclusion.
The retail electronic payments system ignited by M-Pesa, commonly labelled mobile
money, has worked as a transactions platform bringing the unbanked into the banking
system. The platform supported the evolution of national retail payments; positioned
banks as a platform to manage micro accounts through virtual savings products; and
enabled the evolution of virtual credit markets, micro insurance, and investments
in government securities. It has also been useful in tracking fraudulent flows into
personal accounts.
63
Third, digitization has enabled fintechs to develop innovative business models with
payments solutions or platforms sitting in commercial banks. Products such as
M-Akiba for micro investors in government securities, M-Kopa for solar energy supply,
One Acre Fund in agriculture, and many other products are making a difference outside
the financial sector. Once digitization has taken root in an economy, it allows for
sustainable business models to be developed and launched on the digital platform
to support other sectors of the economy.
Finally, digitization is now driving fiscal policy designs, revenue administration, and
public finance management. It is reducing leakages in revenue administration, but
more importantly the digital tax payment platform is an important innovation for
efficiency and transparency. Indeed, digital platforms have revolutionized the way
payments to and from the government are made. In addition, government services
are being digitized across ministries and government agencies.
6.2 Lessons from use of M-Pesa as a tool for

financial inclusion
1. Endogenous demand for regulatory reforms, regulatory capacity and regulatory
technology and, in addition, endogenous demand to complete the financial
infrastructure: Information capital (credit reference bureaus; deposit insurance;
competition policy; consumer protection/financial literacy, etc) are the institutions
that support the market and protect the markets. The Central Bank of Kenya with
the support of the Kenya Bankers Association and financial support from FSD(K)
was able to set up a credit information sharing project that gave rise to Credit
Reference Bureaus (CRBs) that are supervised by the Central Bank. Banks and
individual customers can access information from the CRBs and can also cross
reference their customers with CRB information. Banks, microfinance institutions,
insurances, pension funds and utility companies provide information to CRBs. The
Central Bank considered this as a development of information capital that would
support market growth and deepening through confidence that information on
market participants was available. Also, transactions and savings data can be used
to generate credit scores for use as the basis to evaluate and price micro credit. The
celebrated M-Shwari type of products (also KCB Pesa, and M-Pawa in Tanzania)
have been the natural development from the virtual banking system. The ability to
change the collateral technology that has been a major barrier to affordable credit
and financial sector growth in many African countries opens avenues for further
financial inclusion and household investments. The second intervention was the
need to enhance the capacity and the capability of the Deposit Protection Fund
(now Kenya Deposit Insurance Corporation - KDIC). The deposit insurance covers
up to Ksh 100,000 (US$ 1,100) and covers over 90% of total accounts in the financial
sector. This has become important in efforts to to promote correspondent banking,
cross-border payments, international remittances and Kenya’s regional banking
64
expansion. Regulatory reforms and regulatory capacity has encouraged further

innovation for mass financial inclusion in Kenya. Finally, financial inclusion has
supported a better AML/CFT regime.
2. Banks’ heavy investments on the technological platform: Financial inclusion has

allowed banks to develop capacity to grow and to serve their market niches –
strong banks can weather shocks and roll out competitive products for their
market niches.
3. Strong banks have emerged – branch outlets, bank deposits and loans accounts
have increased. Kenya banks cover the EAC region, thus widening the market.
Between 2007 and 2019, branch networks of Kenyan banks expanded from about
575 to 1,490 branches. The rural branches have expanded faster from about 170
in 2007 to 660 in 2016 and microfinances have also equally increased. In addition,
Kenyan banks have expanded to the Eastern Africa region, with over 310 branch
outlets. The customer base has also increased. Deposit accounts have increased
from about 4.72 million accounts in 2007 to over 55 million accounts in 2018. This
has provided banks with a large deposit base and capacity for growth.
4. Payments technology has now developed; it is efficient, transparent and effective –

has covered other markets/sectors – insurance, capital markets, social protection.
There are different actors at different market segments: microfinance, SACCOs
and Agency banking. M-Pesa has been useful in other financial markets such
as insurance, capital markets and pensions and government’s social protection
programme.
5. Governments targeted intervention on social protection programmes. An efficient

and effective payments platform has been useful for government’s targeted social
protection for poor households and physically disadvantaged persons for financial
transfers.
6. M-Pesa has led to a vibrancy of the financial market cutting across all sub-sectors
and improved transactions at all levels including the informal sector. M-Pesa has
been an important tool for revolutionizing the payments system in the country;
it has been a catalyst and a driver for financial inclusion in Kenya.
7. M-Pesa has improved the environment for monetary policy: Financial inclusion and
financial market development have improved the monetary policy environment.
Declining velocity – is an indication of financial depth and rising multiplier - an
indication of financial innovation. Currency outside the banking sector as a
ratio of broad money has declined, which is a signal for less money being held
in “unsafe’’ places. These outcomes then portend a challenge to the current
monetary policy framework. Velocity movements may imply unstable money
65
demand. The relationship between reserve money and broad money is unstable
and unpredictable, therefore moving from monetary targeting to price-based
monetary policy framework.
6.3 Replication of regulation and other approaches

This regulatory approach by Kenya has been replicated elsewhere. For instance, in
Tanzania, the regulator made a progressive decision to let regulation follow innovation
and support financial inclusion while managing risks. Castri and Gidvani (2014) point
out that by engaging closely with MNOs (and their respective partner banks), the Bank
of Tanzania (BOT) has been able to offer the private sector a degree of freedom in rolling
out new products, responding with sufficient safeguards where necessary. By 2010,
Castri and Gidvani (2014) show that the market had reached a certain level of maturity,
with four providers and more than 10 million registered mobile money customers,
and the BOT had progressively increased its operational knowledge of mobile money
and was now in a position to draft regulations that would provide more legal certainty
to providers. The BOT also had to ensure that the regulatory arrangements were in
compliance with supporting laws and regulations, such as the AML/CFT regime. As
the mobile financial services market evolved, continued emphasis on the “test and
learn” approach needed a continuous revision to shift to other regulatory approaches
such as innovation offices, regulatory sandboxes and RegTech. These are interesting
examples that require expounding on briefly:
Innovation offices
Innovation offices are functional units that play a key role in facilitating regulator-innovator
engagement. UNSGSA FinTech Working Group and CCAF (2019) point out that innovation
offices engage with and provide regulatory clarification to financial services providers
that seek to offer innovative products and services. They can improve understanding of
technology-enabled financial innovation and support appropriate regulatory responses.
They may reduce regulatory uncertainty and signal a pro-innovation stance, which in turn
encourages inclusive FinTech. UNSGSA FinTech Working Group and CCAF (2019) note
that for innovation offices to be effective, there should be early and close engagement
with innovators for executive buy-in and inter-agency coordination.
Regulatory sandboxes
A regulatory sandbox is a regulatory approach typically summarized in writing

and published, and which allows live, time-bound testing of innovations under a
regulator’s oversight. It consists of a set of rules that allow innovators to test their
products/business models in live environment without following some or all legal
requirements, subject to predefined restrictions. UNSGSA FinTech Working Group
and CCAF (2019) explain that to date, at least two discernible models have emerged:
66
(i) product testing sandboxes; and (ii) policy testing sandboxes. The lines between
the two are not rigid. As pointed out by UNSGSA FinTech Working Group and CCAF
(2019), there is emerging evidence of policy testing within product-focused sandboxes
particularly in the context of thematic cohorts. Sandboxes can help regulators gain a
better understanding of fintech and develop evidence-based regulations that promote
inclusive fintech. A regulatory sandbox brings the cost of innovation down, reduces
barriers to entry, and allows regulators to collect important insights before deciding
if further regulatory action is necessary.
RegTech
RegTech (regulatory technology) is a distinct innovative regulatory initiative. UNSGSA

FinTech Working Group and CCAF (2019) point out that RegTech focuses on how to
monitor and enforce compliance against relevant regulations, and thus can support a
more responsible delivery of innovative financial services, which may directly impact
financial inclusion. It also allows regulators to swiftly respond to market developments,
better protect consumers, and enhance institutional supervision. UNSGSA FinTech
Working Group and CCAF (2019) argue that in the recent past, regulators themselves
have begun to consider RegTech as a tool to help keep up with the substantial changes
in financial services marketplaces, with RegTech being considered as two distinct but
complementary branches; that is, compliance technology (CompTech) and supervisory
technology (SupTech).
6.4 The challenges
The digitization of economies comes with many benefits of strengthening government

capacity but also raises new demands on capacity for executive regulation to address
the challenges. To reap the full range of potential benefits, Kenya will need to ensure a
competitive ecosystem and infrastructure that facilitates entry. An enabling regulatory
environment and robust consumer protection will also be critical.
1. Securing interoperability of retail electronic payments and

transparent pricing
The interoperability14 of Mobile Network Operators (MNOs) and transparency in the

costing of services is still a concern for telecommunication regulators and financial
institutions. Kenya’s National Payment System Act of 2011 requires payment service
providers to use systems capable of becoming interoperable with other payment
systems in the country and internationally. However, there are large imbalances in the
shares of a mobile market that has almost reached saturation coverage. As of June
14 Interoperability can broadly be described as the interconnection of mobile money services either
between providers or with external parties (Mazer and Rowan, 2016).
67
2020, the number of mobile subscriptions stood at 57.03 million corresponding to a

mobile penetration rate of 119.8%. Safaricom is the dominant MNO in Kenya with a
market share of 64.2%, followed by Airtel with a market share of 26.8% and Telkom
Kenya (previously Orange Kenya) with a market share of 6.0% and finally Equitel with
subscription market share of 3.0%. The fact that Safaricom continues to dominate the
market creates a fundamental conflict of interest since all other financial institutions
in Kenya remain customers of, and competitors to, one dominant MNO.
The conflict of interest derives from the fact that Safaricom controls access to a
large proportion of the mobile network and has an incentive to restrict access to
competitors. Mazer and Rowan (2016) point out that the presence of a dominant MNO
leaves third-party providers with no other option to reach the majority of the market
than to go through this MNO. This implies little incentive for the MNO to drive down
the price of unstructured supplementary service data (USSD) technology, which is the
dominant front-end technology used in the deployment of mobile banking services
in Kenya. The dominant MNO has considerable power to set prices in the market and
to control competition by providing or restricting access. Without interoperability,
consumer freedom to switch to other networks is limited; accounts are effectively
restricted to those on the same MNO network, Agents are not likely to function as
Agents for multiple MFS providers, and access to MFS channels by third parties is
restricted (Mazer and Rowan, 2016). In a near-saturated market especially, there
will be less willingness on the part of Safaricom, as the dominant MNO, to extend
interoperability since it may have more to gain by protecting its share of the pie rather
than expanding the size of it. Paradoxically, the concentration in market share across
mobile financial service providers may also create less demand for interoperability
from consumers, since most of their peers will use the same provider.
However, these concerns, real as they are, seem to oversimplify the debate on
interoperability in Kenya. It appears that most researchers considering this issue
ignore four important elements of the market structure that developed soon after
M-Pesa’s success.
• First, Safaricom has built a network infrastructure for connectivity that is larger than
any other MNO and which it would be very expensive for other MNOs to replicate.
A viable solution could be a market arrangement for leasing/renting the installed
telcom infrastructure by Safaricom to other MNOs and Mobile Virtual Network
Operators (MVNOs)s in a given locality based on utilization, as an important avenue
for increasing competition and moving interoperability to a higher level.
• Second, the Agent management model has been structured around Safaricom
activities. The regulators imposed supervisory responsibility on Safaricom for its
Agents, and the resulting master Agents model, described previously, is largely
responsible for M-Pesa’s success. Even when the Agents’ exclusivity clause was
removed, other MNOs did not move to engage with the existing Agent network, which
68
could have created fertile grounds for competition and demands for interoperability.
Instead, they have continued with their relatively small number of Agents. The
argument here is that interoperability is supposed to increase the market size,
lower unit costs and enhance competition and efficiency, leaving market shares to
depend on the products and services rolled out by the different MNOs.
• Third, it is much easier for businesses and financial institutions to integrate with
M-Pesa as a payments platform. This is in part due to M-Pesa capturing the first-
mover advantage, but it also reflects M-Pesa’s continued heavy investments in
infrastructure, both connectivity and payments platforms, which other MNOs have
not been able to replicate. For the debate on interoperability to move to the next
level, the market structure in Kenya need to be appreciated. This is why progress
probably implies developing arrangements for Safaricom to lease or rent out its
existing infrastructure to other MNOs and MVNOs.
• Fourth, Safaricom also leads in another important dimension. It has rolled out
a range of complementary products that other MNOs have failed to develop,
including support for internet connectivity, home fibre and CCTV security
networks, even for use by the government. The lack of product diversity by other
MNOs and MVNOs has restricted their coverage and reach. Getting to a competitive
market will not be a simple process.
In May 2018, interoperable transactions were formally launched. So far, these cross-
network transactions are not in real-time like those of M-Pesa, and their unit cost is
much higher. However, these challenges can be overcome if all MNOs invest in real-
time transaction platforms, which would reduce the unit costs of transactions.
2. Connectivity across the country
According to the World Bank (2016), there are some 4 billion people worldwide without
internet access, about 2 billion who do not use mobile phones, and almost half a
billion who live outside areas with a mobile signal. Currently, there are still a number
of mobile users in Kenya on 2G technologies, with many more on 3G technologies and
4G technologies. 5G technology is yet to be rolled out. According to Sanni (2017), as of
2016, Kenya’s mobile subscribers market penetration was at 54%, its mobile internet
subscribers’ penetration was at 52%, and its proportion of 3G connections was at 22%
while its proportion of 4G connections was only 2%.15
The GSMA’s Mobile Connectivity Index measures individual countries’ mobile internet
penetration based on four key enablers for mobile internet adoption: infrastructure,
affordability, consumer readiness, and content. Sanni (2017) argues that to enable
15 Most of these indicators compare favourably with Sub-Saharan Africa’s averages: 43% for mobile
subscribers’ market penetration, 26% for mobile internet subscriber penetration, 31% for
proportion of 3G connections, and 2% for proportion of 4G connections.
69
greater connectivity, Kenya needs to address a number of these enablers, particularly

around network quality, affordability for low-income users, digital skills, and the
creation of content in more local languages. If not adequately addressed, the exclusion
resulting from these connectivity challenges is likely to further widen the digital
divide in Kenya, which is a great concern to the adoption and uptake of e-government
services. Power is another constraint; connection to the national electricity grid is
limited, especially in rural areas, and there are frequent power blackouts across the
country, including for Huduma centres.
Recognizing the gap, the Kenyan Government has made efforts to improve connectivity
nationwide. As at June 2016, the Government had laid 6,000 km of national fibre
optic backbone across all the 47 counties. The installation of internet connectivity
equipment has been completed in 29 counties. Installation works were in progress
in 12 other counties while five counties were awaiting approval of design. The
Government intends to engage private sector players to conclude the installation of
the national fibre optic backbone infrastructure. There are also proposals for a project
to develop constituency digital innovation hubs to support entrepreneurs, including
access to free Wi-Fi in all the 290 constituencies. Improved connectivity across the
country would further reduce the cost of doing business, bridge the digital divide, and
enhance the roll-out and use of innovative products on digital platforms.
3. Regulatory challenges
Several regulatory challenges have arisen from the use of digital platforms in general and
digital financial services in particular. First, preventing cybercrime remains a challenge for
Kenyan regulators, especially in terms of capacity for adequate office-level surveillance.
A 2016 cybercrime and cybersecurity study by Symantec ranked Kenya as the prime
source state in Sub-Saharan Africa as measured by cyber-attack numbers, malware,
spam, and phishing hosts (Didenko, 2017). Kenya lacks a comprehensive legal framework
for addressing cybercrime. This will be addressed through legislative reforms that are
underway following the publication of the Computer and Cybercrimes Bill in June 2017.
The second challenge is the lack of specific crowdfunding regulations in Kenya or in

generating new regulations to cope with the innovations and new business designs. A
number of regulators have authority to regulate various forms of crowdfunding, including
the Central Bank of Kenya, the Capital Markets Authority, and the Communications
Authority (Didenko, 2017). A recent study by the Cambridge Centre for Alternative
Finance identified several statutes and regulations that might apply to the operation
of crowdfunding platforms, depending on their mode of operation. These include
the National Payment Systems Act, the Money Remittance Regulations, the Kenya
Information and Communications Regulations, the Microfinance Act, the Proceeds of
Crime and Anti-Money Laundering Act, the Capital Markets Act, the Banking Act, and the
Public Offer Regulations (Didenko, 2017). The Public Fundraising Appeals Bill proposes
additional licensing requirements in connection with fundraising systems.
70
References
Castri S. and Gidvani L. 2014. Enabling mobile money policies in Tanzania: A ‘Test and Learn’
approach to enabling market-led digital financial services. DO-10.2139/SSRN.2425340.
Central Bank of Kenya, https://www.centralbank.go.ke/reports/bank-supervision-and-
banking-sector-reports/, accessed in October 2020.
Central Bank of Kenya, database: https://www.centralbank.go.ke/statistics/, accessed in
September 2020.
Communications Authority of Kenya – CAK. 2020. Fourth quarter sector statistics for the financial
year 2019/2020. https://ca.go.ke/document/sector-statistics-report-q4-2019-2020/,
accessed in October 2020.
Didenko A. 2017. Regulatory challenges underlying fintech in Kenya and South Africa. British
Institute of International and Comparative Law, Bingham centre, The Rule of Law. https://
www.biicl.org/documents/1814_regulation_of_fintech_in_kenya_and_south_africa_v_1.pdf.
FSD Kenya and Central Bank of Kenya. 2019. FinAccess National Household Survey, 2019:
Access, usage, quality and impact. Nairobi: FSD Kenya and Central Bank of Kenya.
Gai, K., Qiu M. and Liu, M. 2018. Privacy-preserving access control using dynamic programming
in fog computing. IEEE 4 th International Conference on Big Data Security on Cloud
(BigDataSecurity), IEEE International Conference on High Performance and Smart
Computing, (HPSC) and IEEE International Conference on Intelligent Data and Security
(IDS), IEEE, New York.
Greenacre, J. 2019. Regulating mobile money: A functional approach. Blavatnik School of
Government, Oxford University.
Mazer, R. and Rowan, P. 2016. “Competition in mobile financial services: Lessons from Kenya
and Tanzania”. The African Journal of Information and Communication, Vol. 17: 39–59.
Mwega, F.M. 2014. Financial regulation in Kenya: Balancing inclusive growth with financial
stability. Working Paper 407. London: Overseas Development Institute.
Muthiora B. 2015. Enabling mobile money policies in Kenya: Fostering a digital financial
revolution, GSMA, mobile money for the unbanked. London, United Kingdom.
Ndungu, N. (1994) “A Monetarist Model of Inflation: Kenyan Case”, African Development Review,
Vol. 6 No. 2, December 1994.
Ndung’u N. 2018. Next steps for the digital revolution in Africa: Inclusive growth and job creation
lessons from Kenya. Brookings Working Paper 20. https://www.brookings.edu/research/
next-steps-for-the-digital-revolution-in-africa/.
Sanni, S. 2017. Socio-economic impact of mobile broadband in Kenya. London: GSMA. Available
at: https://www.gsma.com/spectrum/wp-content/uploads/.
71
UNSGSA FinTech Working Group and CCAF. 2019. Early lessons on regulatory innovations to
enable inclusive fintech: Innovation offices, regulatory sandboxes, and RegTech. New York
and Cambridge, UK: Office of the UNSGSA and CCAF.
Vaughan, P., Fengler, W., Joseph, M. (2013). Scaling-up through disruptive business models.
The inside story of mobile money in Kenya. In Chandy, L., Hosono, A., Kharas, H., Linn, J.
(Eds.), Getting to scale: How to bring development solutions to millions of poor people
(pp. 189–219). Brookings Institution Press.
World Bank. 2016. Digital Dividends: World Development Report 2016. Washington DC: World
Bank.

72
ANNEX 1
the ‘audit’
report on M-Pesa
executive summary
The objective of this brief is to shed light on:
• How M-Pesa, an innovative money transfer service, has developed in Kenya

• How it is providing services to the unbanked in Kenya both in the urban and rural
• Further how the Central Bank of Kenya continues to play an oversight role to
ensure its safety and efficiency, as part of the national payments system role
• Finally, to discuss any risks it poses and how these risks are mitigated
M-Pesa is an electronic money transfer product that enables users to store value on
their mobile phone or mobile account in the form of electronic currency that can
be used for multiple purposes, including transfers to other users and conversion to
and from cash. Electronic money services contribute to the development of mobile
commerce and have an overall positive impact on the economy through increased
velocity of money (transmission) and swift settlement of transactions. Moreover, the
service enables unbanked members of society to access financial services in areas
unserved by formal banking services.
The brief distinguishes M-Pesa money transfer service from banking business as
provided for under Section 2(1) of the Banking Act. In the M-Pesa model, money is
held in a trust account at the Commercial Bank of Africa. Since money in the trust
account is not under the control of Safaricom and further cannot be employed for
such purposes as lending, investing or in any other manner for the account and at
the risk of Safaricom as per Section 2(1) of the Banking Act, M-Pesa is not equivalent
to a bank account.
Though technological developments such as ATMs, internet, and mobile phones all
have high potential for money transfer services, mobile phones have a higher degree
of penetration unlike the ATMs and internet, which have limited coverage due to their
dependability on electricity and telephone (fixed) lines.
73
The trust deed that provides legal protection for the money in the trust account is
provided for in the trustee deed. Various legal instruments pertaining to this service
including the trustee deed have been presented to the Central Bank and reviewed
accordingly. Further to this, funds in the trust account deposited in the Commercial
Bank of Africa are regulated by the Central Bank of Kenya under the Banking Act.
Prior to launching the M-Pesa service, Safaricom requested CBK and CCK to authorize
the money transfer service. In evaluating this, both institutions had to ensure safety
and adequacy of consumer protection. This is still safeguarded as a goal and also to
safeguard the credibility of the service.
Legal and regulatory efforts undertaken by both the Central Bank and National
Treasury regarding safety, efficiency, and effectiveness of payment systems in the
country are highlighted. These include the review of the Central Bank of Kenya
Act in the year 2003 to include Section 4A 1(d), the drafting and forwarding to the
Attorney General of the National Payments System Bill, and the development of the
Oversight Policy Framework document on payments systems in the country. These
developments are also benchmarked against International Standards recommended
by the Bank for International Settlements (BIS).
M-Pesa is still a low value transfer service and has penetrated all parts of the economy
and serves both the banked and unbanked population. But its critical role is to transfer
funds. The number of customers using the service has increased from 876,000 in
October 2007 to 3.7 million in August 2008. In the same period, transactions have
increased from Ksh 2.830 billion in October 2007 to Ksh 16.756 Billion in August 2008.
But the average value per transaction has changed marginally from Ksh 2,951 average
per transaction in October 2007 to Ksh 2,642 in August 2008 and Ksh 2,916 as the
average the whole period the M-Pesa service has been in operation.
1. introdcution
Kenya is among a number of countries where financial services are starting to be offered
by mobile network operators. There is considerable interest in the development of these
services since they offer the prospect of providing services to people who presently do
not have bank accounts. Branchless banking through retail Agents in Kenya is made
possible through the information and communication technologies that customers,
retail Agents and mobile network operators use to record and communicate transaction
details. A very wide range of outlets are used to offer payment services, which include
supermarkets, petrol stations, airtime sellers, courier companies and fertilizer merchants
and post offices. Branchless banking through retail Agents may be far more convenient
and efficient for poor customers than a bank branch, and this has resulted in extending
financial services to the unbanked and marginalized population.
74
Kenya has experienced quite a significant growth in mobile adoption in the last
few years. For example, currently there are 15.2 million mobile phone subscribers
compared to 15,000 seven years ago. Results from the DfID-funded FinAccess study
(August/September 2006) showed that over half of the adult population in Kenya
either own their own phone or have access to one through a friend or family member.
In Kenya, while over 55% of the population has access to mobile phones, only 19%
of the population is banked. It does follow then that the market is large, that can be
tapped by both banks and mobile phones.
The M-Pesa product which was developed by Vodafone and which is operated by
their subsidiary and largest Mobile Operator-Safaricom was launched in March 2007
following an initial pilot. This pilot was co-funded by Vodafone and the DfID Financial
Deepening Challenge Fund. This brief addresses the issues of M-Pesa and other related
services in order to show their workings and related risks, and how these risks are
mitigated.

the m-pesa product
Kenyans are using the M-Pesa service to transfer money safely, efficiently and effectively.
They use it for paying field staff their allowances and expenses so that they do not need
to travel to the Head Offices for payment, sending a long haul truck driver money for
spare parts, sending money to family members for consumer purchases, school fees
payment, sending pocket money to students in schools, and sending emergency medical
payments among other purposes.
A taxi driver wishing to be offered prepaid services due to security reasons could
request for payment via M-Pesa service. In many instances today, Kenyans traveling
up-country deposit cash before the start of the journey to pick it up upon arrival
to their destination thus avoiding the risk of loss through theft or robbery that has
increased in Kenyan highways today. The M-Pesa service is therefore not only safe,
secure, and fast but meets the needs of many Kenyans thus its popularity across
Kenyans of all divide.
m-pesa is a money transfer service not a bank account
According to Section 2(1) of the Banking Act, “banking business” means:
a) the accepting from members of the public of money on deposit repayable on

demand or at the expiry of a fixed period or after notice;
b) the accepting from members of the public of money on current account and
payment on and acceptance of cheques; and
75
c) the employing of money held on deposit or on current account, or any part of the
money, by lending, investment or in any other manner for the account and at the
risk of the person so employing the money.
Section 2(1) (c) above, empowers banks to take risk by employing monies received
as deposits by lending or investing such monies. This banking business is the one
regulated under the Banking Act. It is within this same mandate that banks under
instructions by their customers as per Section 2(1) (b) transfer money.
In a national survey on Access to Financial Services in Kenya, it was established that

prior to the launch of M-Pesa, local money transfers were effected through various
means and include through family/friend (42%), bus or matatu (20%), money transfer
services (7%), post office (18%), directly into bank account (8%), cheque (3%), and
paid into someone else’s account (2%). It is therefore established that money transfer
services in Kenya have been inefficient and relatively by unsafe means. Technological
developments including ATMs, internet, and mobile phones have a high potential to
transfer money more efficiently, effectively, and safely with the respective transaction
charges greatly reduced through appropriate use of technology. But while this is the
case, both ATMs and internet are electricity- and telephone lines-dependent and are
therefore only available in limited areas where electricity and telephone services have
been extended. Safaricom’s M-Pesa service, however, is mobile phone-dependent
and has inherent high penetration capabilities to urban, rural, and remote areas of
the country. In view of the foregoing, therefore, the Central Bank of Kenya and other
regulatory bodies have a challenge to provide an enabling environment for delivery
of these technologically driven innovative money transfer services in order to deepen
access to financial services in Kenya.
In M-Pesa, money collected by Agents is deposited in a trust account at the Commercial

Bank of Africa. This trust account provides the legal protection for the beneficiaries.
The money in this trust account is not under the control of Safaricom and cannot
be employed for purposes such as lending, investing or in any other manner for the
account and at the risk of Safaricom as per Section 2(1) of the Banking Act. Legal
protection of the money in the trust account is provided for in the trustee deed.
Various legal instruments pertaining to this service including the trustee deed have
been presented to the Central Bank and reviewed accordingly. Further to this, funds
in the trust account deposited in the Commercial Bank of Africa are regulated by the
Central Bank of Kenya under the Banking Act.
76
trust deed
“Between M-Pesa holding company (trustee)

and M-Pesa participants”
The Trustee holds funds on behalf of all M-Pesa System Participants under a
Declaration of Trust (the Trust Deed). Highlights of the Trust Deed are:
o The Trustee holds all amounts which constitute the Trust Fund on trust for the
System Participants.
o The beneficial entitlement of each System Participant to the Trust Fund at any time
shall be to such amount of the Trust Fund in conventional money as is equal to the
amount of E-Money in the M-Pesa Account of such System Participant at such time.
o Safaricom is entitled to levy certain charges on System Participants for the

operation of the Service. Where it does so, the M-Pesa Account of the relevant
System Participant will be debited by the amount in E-Money of the relevant charge
and a M-Pesa Account of Safaricom shall be credited with the relevant amount.
o The amounts constituting the Trust Fund shall be held by the Trustee in Commercial
Bank of Africa.
o Safaricom undertakes to the Trustee and to the System Participants that it will not
issue any new E-Money other than in return for an equal amount in conventional
money being paid to and received by the Trustee.
o Safaricom shall also not effect any transfer of any E-Money from any M-Pesa
Account of an amount which exceeds the credit balance of E-Money in the relevant
M-Pesa Account.

what risks does m-pesa face? how are they mitigated
Risks in M-Pesa money transfer service are similar to those that attach to payments
systems worldwide. These include:
1.1 Credit risk – the risk that a counterparty will not meet an obligation for full value
either when due, or at any time thereafter
Mitigation: This is mitigated by ensuring that Agents prepay before offering

services to customers. These Agents operate under an agreement that is enforced
by Safaricom.
77
1.2 Liquidity risk – the risk that a counterparty will not settle an obligation for full
value when due, but at some time thereafter
Mitigation: Safaricom employs stringent vetting criteria which includes entities

that are financially sound. As at September 2008, the Agents were distributed as
follows; Commercial banks (3.9%), Postal Corporation (2.2%), Supermarkets (1.5%),
Petrol stations (1.3%), Shops (24.5%), Airtime sellers (60.7%), SACCOS (3.2%), Courier
companies (0.9%), and Chemists (0.9%).
In the recent past and following discussions with the Central Bank of Kenya, the
Agents include commercial banks and ATMs.
1.3 Operational risk – the risk that hardware or software problems, or human error,
or malicious attack will cause a system to break down or malfunction giving rise
to financial exposures and possible losses
Mitigation: Safaricom is part of the Vodafone group, an international and reputable

multinational in the provision of mobile technology. The M-Pesa product benefits
from research and development of Vodafone. Operational issues are regulated by
the CCK.
Central Bank of Kenya receives reports on operational issues on a monthly basis. The
Bank has emphasized the need for Safaricom to ensure adequate disaster recovery
and business continuity arrangements.
1.4 Settlement risk – The risk that the flow of funds between transacting parties fails
or delays owing to credit, liquidity and operational risk or use of a risky settlement
medium which does not coordinate delivery and payment.
Mitigation: Agents of Safaricom settle through the trust account at the Commercial
Bank of Africa and since M-Pesa is a high volume low value retail payment system,
settlement in a sound commercial bank is deemed adequate. In authorizing the
M-Pesa service, and bearing in mind settlement risks, Central Bank of Kenya placed
a maximum limit of Ksh 50,000 per M-Pesa account per day and a transaction limit
of Ksh 35,000 per transaction.
1.5 Foreign Exchange Settlement Risk (Herstat risk) – the risk that one party to a
foreign exchange transaction does not receive the foreign currency it paid for.
Mitigation: Safaricom has proposed a foreign remittance service which would

require mitigation of this risk. Central Bank of Kenya is reviewing documents
supplied by Safaricom. The Bank has also advised Safaricom on the requirement
of this service in line with the provisions of the Central Bank of Kenya Act.
78
1.6 Legal risk – the risk that unexpected interpretation of the law or legal uncertainty
will leave the payment system or members with unforeseen financial exposures
and possible losses
Mitigation: Safaricom as a mobile service provider is licensed and regulated by the

CCK and is therefore under the provisions of the Kenya Communications Act 1998.
As indicated above, legal relationships with respect to money transfer are provided
for in various agreements which are reviewed by the Central Bank.This innovative
service is a value added on the licensed mobile services. As a money transfer service,
use is made of the existing oversight mandate of the Central Bank of Kenya. This will
further be enhanced with enactment of the proposed National Payments System
Bill, which has been forwarded to Treasury and the Attorney General.
Anti-Money Laundering (AML) and Combating Financing of Terrorism (CFT):

From the outset and even before authorization, Safaricom was advised on the
requirements of AML/CFT. In this respect, therefore, Safaricom has continued to
mitigate this risk through legal instruments, training of Agents, system monitoring
for suspicious transactions, and enforcement. In addition, Safaricom observes AML/
CFT policy requirements of Vodafone. Just like any other institution, Safaricom
awaits the enactment of the AML Bill.
2. legal and regulatory framework

Oversight of Payments System in Kenya
Following the request by Safaricom to launch this innovative money transfer service
in early 2007, the Central Bank of Kenya established a committee which worked
closely with payment system experts to facilitate mobile banking in Kenya. A number
of critical issues were evaluated, including: the security of the system and other
operational aspects, the registration procedures, the systems audit trail, and consumer
protection issues. On its part, the Central Bank was more concerned with the need
to mitigate various risks that would compromise safety, efficiency, and effectiveness
of the payment system.
Oversight of mobile payments is premised on the provisions of the Central

Bank of Kenya Act whose mandate was expanded in 2003 to include Section
4A 1(d), which inter alia mandates the Bank to formulate such policies as best
promote the establishment, regulation and supervision of efficient and effective
payment, clearing and settlement systems. Pursuant to this mandate and in a
bid to adequately cater for the modernization of the payments system, the Bank
has formulated and proposed the enactment of the National Payments System
Act. While the proposed Act will address low value retail payment systems such
as M-Pesa, it will also provide for Large Value Payment Systems for inter-bank
79
payments such as the Kenya Electronic Payments and Settlement System (KEPSS),
which is a Systemically Important Payment System (SIPS).
Further to this mandate and in order to operationalize the oversight function, the
Bank has developed an Oversight Policy Framework document on payments system
in Kenya, which is available at the Bank’s website, www.centralbank.go.ke
Prior to the launch of the M-Pesa service and in a meeting held on 16th February
2007 at the Treasury, the following issues were discussed by the Treasury, Ministry of
Information and Communications, Central Bank of Kenya, and Safaricom:
• Given that M-Pesa is an innovative e-based money transfer service, Basel Core
Principles on payment, clearing and settlement systems and the risk management
principles for e-banking will govern its operations;
• The need to clearly define the relationship between Safaricom and Commercial
Bank of Africa (CBA), the providers of settlement mechanism for the product;
• The need to clearly define the Agency relationship between Safaricom and M-Pesa
Agents;
• The relationship between Safaricom and its primary regulators, CCK, with regard
to M-Pesa as a value adding product and tax implications, if any;
• The role of Postal Corporation of Kenya (PCK) as a party interested in establishing

agency relationships with Safaricom on the M-Pesa product.
The meeting recommended the following:
• There is need for Safaricom to familiarize itself with the requirements set out in
the Basel regulatory instruments;
• Bearing in mind that the Central Bank of Kenya Act has mandated Central Bank
to have oversight over payment and settlement systems, Safaricom should
initially facilitate CBK’s access to information on the M-Pesa product, pending
the enactment of specific and comprehensive payment system legislation;
• In the meantime, transparent and clear mandate of the settlement and clearing
banks, e.g. CBA, should be upheld;
80
• The relationship between Safaricom and its primary regulator CCK, with regard to
M-Pesa as a value adding product and tax implication, for airtime is maintained
on a compliant basis.
• Safaricom should ensure to identify and mitigate risks associated with the product,
including settlement risk, infrastructure risk, security risk, and data control risk.
As a minimum, adequate audit trail should be maintained.
• The role of PCK as a party interested in establishing agency relationship with

Safaricom on the M-Pesa is encouraged due to the wide network of PCK.
Subsequently, the firm was authorized to provide payment services and required to
mitigate risks thereto. CBK has continued to oversee the service in line with these
recommendations. M-Pesa is not an alternative bank account and it has remained a
low value high volume money transfer service as was envisaged since its inception.
The use of the mobile phone platform for delivery of financial services is not limited
to M-Pesa. Safaricom is therefore not the only mobile phone money transfer service
provider in Kenya as commercial banks are using it to deliver traditional banking
products, which include: account inquiry (balance inquiry and mini-statement),
funds transfer, bills payment, mobile recharge, and other requests (cheque book and
statement request) at fees ranging from Ksh 20 to Ksh 100.
In addition, other mobile companies in Kenya such as Zain, formally Celtel, have
been developing money transfer services. Celtel had developed the Sokotele product
which Zain is proposing to upgrade to ZAP. Econet has proposed to introduce a similar
product with product name Obopay.
CBK has been evaluating these products to ensure that they meet safety and efficiency
requirements prior to being launched.
M-Pesa Deposits and Withdrawals
Data on M-Pesa deposits and withdrawals for the period October 2007 to August
2008 is provided in Table A1 while that relating to amounts transferred and their
corresponding transactions for the period March 2007 to August 2008 is provided
in Table A2 (below). The low average net deposit per customer of Ksh 203 confirms
that M-Pesa is a money transfer service while the low average transfer amount per
transaction of Ksh 2,916 indicates that M-Pesa is a low value high volume retail
payment system.

81
Table A1: Deposits and withdrawals

Total Deposits Total Withdrawals Net Deposits (KSh) No. of Net
(KSh) (KSh) Customers Deposit per
customer
(Ksh)
Oct 07 1,475,563,981 1,353,985,998 121,577,983 875,962 139
Nov 07 1,854,171,597 1,660,781,690 193,389,907 1,133,202 171
Dec 07 2,003,771,945 1,766,498,444 237,273,501 1,345,269 176
Jan 08 2,219,896,512 1,839,143,333 380,753,179 1,589,100 240
Feb 08 2,760,217,824 2,459,572,110 300,645,715 1,821,533 165
Mar 08 3,565,425,759 3,182,026,434 383,399,325 2,075,527 185
Apr 08 4,427,816,102 3,961,822,776 465,993,326 2,373,455 196
May 08 5,777,370,485 5,126,784,590 650,585,895 2,718,127 239
Jun 08 5,768,474,049 5,148,605,901 619,868,148 3,038,523 204
Jul 08 7,425,318,485 6,591,776,408 833,542,077 3,367,192 248
Aug 08 8,872,580,175 7,883,680,776 988,899,399 3,726,175 265
Average net deposit per customer 203
M-Pesa is still a low value transfer service and has penetrated all parts of the economy
and serves both the banked and unbanked population. But its critical role is to transfer
funds. The number of customers using the service has increased from 876,000 in
October 2007 to 3.7 million in August 2008. In the same period, transactions have
increased from Ksh 2.830 billion in October 2007 to Ksh 16.756 billion in August
2008. But the average value per transaction has changed marginally from Ksh 2,951
average per transaction in October 2007 to Ksh 2,642 in August 2008 and Ksh 2,916
as the average the whole period the M-Pesa service has been in operation. Similarly,
the number of transactions has increased from 21,714 in May 2007 to 6,342,413 by
August 2008 (Table A2). This tremendous growth has also been matched by the growth
in transactions value.
82
Table A2: Monthly transactions details

Total value of Total transactions Average value per
transactions (Ksh transaction (Ksh)
million)
Mar 07 64 21,714 2,965
Apr 07 221 69,740 3,167
May 07 484 149,986 3,225
Jun 07 720 233,661 3,082
Jul 07 1,065 354,298 3,007
Aug 07 1,580 516,239 3,060
Sep 07 2,070 669,689 3,091
Oct 07 2,830 958,908 2,951
Nov 07 3,515 1,221,742 2,877
Dec 07 3,770 1,274,098 2,959
Jan 08 4,059 1,346,827 3,014
Feb 08 5,220 1,739,903 3,000
Mar 08 6,747 2,397,498 2,814
Apr 08 8,390 3,072,888 2,730
May 08 10,904 4,021,265 2,712
Jun 08 10,917 4,201,440 2,598
Jul 08 14,017 5,381,073 2,600
Aug 08 16,756 6,342,413 2,642
Average transfer amount per transactions 2,916
This has worked in several distinct ways that can be summarized as: financial inclusion
outcomes; the improvements of AML/CFT especially because DFS has reduced
drastically informality of financial market segments; strong banks have evolved in
Kenya as a results of improvement in the intermediation process both in quantum
and efficiency; collateral process and collateral technology has changed and so
has development of information capital an finally the terrain for monetary policy
effectiveness has improved.
83
A Digital Financial Services Revolution in Kenya: The M-Pesa Case Study
This case study aims to contribute to discussion around appropriate regulatory and policy
issues of mobile money by focusing on actions that the Central Bank of Kenya took to help
M-Pesa grow in a sustainable way. The volume provides a unique, personal perspective of
the development of M-Pesa and discusses the need to balance innovation and systemic risk
while promoting competition. Competition itself can be a difficult concept when regulating
an innovation being launched by a monopoly into a market for financial services containing
dominant institutions.
Taken together, the chapters characterize generations of M-Pesa, covering initial growth, bank
linkages, digital credit, international remittances and fintech. It further notes the monetary
policy impact of M-Pesa in the velocity of money circulation and the money multiplier in
the money supply process that pushed the changes in monetary policy framework. These
generations demonstrate how the continuing journey of M-Pesa, which has hugely benefited
Kenyans, has facilitated rapid growth in financial technology and stimulated digital government
and digital transformation.
Prof. Njuguna Ndung’u is the Executive Director of the African Economic Research Consortium
(AERC) and the immediate past Governor of the Central Bank of Kenya (CBK), where he
successfully completed the full two 4-year terms, 2007-2015. His tenure saw the rapid rise of
Kenya in the financial inclusion space with global recognition stemming from the pioneering
and transformative M-Pesa. He is an Associate Professor of Economics, University of Nairobi,
and faculty member of the School of Economics since 1987. He has published widely in
international journals as well as chapters in various books on economic policy issues. Prof
Njuguna Ndung’u has also been a member of the Advisory Committee of the Alliance for
Financial Inclusion (AFI) and was its Chair in its formative years in 2009-2012.
ISBN: 978 9966 61 112 3
African Economic Research Consortium

A Digital Financial Services Revolution in Kenya

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Digital Financial Services Revolution in Kenya

Uploaded by

Copyright:

Available Formats

A Digital Financial Services

The M-Pesa Case Study

Established in 1988, African Economic Research Consortium is a premier capacity

Networking – the linking of individuals and institutions in a knowledge-sharing,

A Digital Financial Services Revolution in Kenya: The M-Pesa Case Study

Published by: African Economic Research Consortium

ISBN: 978 9966 61 112 3

© 2021, African Economic Research Consortium

6. Lessons, Challenges and the Future for Digitization................................63

List of abbreviations and acronyms

• Cashing in on digital revolution: “Digitization makes finance accessible, lowers

• Practitioner’s insight: M-Pesa, a success story of digital financial inclusion,

• “Boosting transformational technology: Creating supportive environments for

In addition, I would like to express my deepest appreciation to Jonathan Espie

Prof. Njuguna Ndung’u

2.1 Financial exclusion

Table 2.1: Financial access by category 2006

• Minimum balance requirements for savings accounts;

• Restrictions on the number of withdrawals per period on a savings account;

• Weak legal framework and incomplete financial infrastructure – for example,

• Customers’ information asymmetry on how banks operate/presence of segmented

2.2 Policy changes

The Communication Bill was passed by Parliament and recognized electronic

3.1 Initial experiments

In August 2006, Safaricom approached the CBK to propose a nation-wide launch

3.2 The operation of M-Pesa

Deposits and withdrawals operated in the following way:

Figure 3.1: M-Pesa operations

3.3 Risks and their mitigation

3.3.1 CBK’s authority and resources

3.3.2 General risks

• The Anti-Money Laundering (AML) and Combating Financing of Terrorism (CFT)

Protection of customers’ funds with cash merchants required particular attention

Safaricom addressed liquidity risk at Agents by employing stringent vetting criteria

3.3.3 Protection of customers’ funds through trust account

The CBK’s classification of M-Pesa as a banking product is particularly important

• Safaricom is entitled to levy certain charges on System Participants for the

3.4 Policy considerations

In the meantime, telco-agency model required further clarification and restrictions.

4.1 Initial growth

To improve on management of transactions, speed and risk mitigation, Safaricom

Table 4.1: Monthly transactions

Perhaps one can look at the network of Safaricom Agents as an infrastructure to

a) A network of financial services providers and points of service outside the

b) E-money transfers, payments and settlement network in real time that is

4.2 Initial audit

• It was not a deposit-taking service.

Box 4.1: PRESS RELEASE M-PESA MONEY TRANSFER SERVICES

The Government initiatives to promote Information Communication Technology and

a) Accept from members of the public money or deposit that is repayable on

7) It is also noteworthy that the recently enacted Kenya Communications (Amendment)

MINISTER FOR FINANCE January 6, 2009

5.1 Generation 1: Initial mobile money growth and

Table 5.1: The entry of the MFS providers in the market

Mobile Network Operator cum MFSP Product Entry Date

3. The monetary policy regime depended on monetary policy instruments that

4. The Government’s Kenya Vision 2030 prescribed increased savings and

In preparation for a commercial launch of mobile money, Safaricom sought approval

Box 5.2: Reflections by Michael Joseph on M-Pesa’s critical moments

Source: Adapted from the Author

Source: Central Bank of Kenya

Source: Central Bank of Kenya

Table 5.2: M-Pesa agents, accounts and transactions