Microfinance

Microfinance
Microfinance
Emerging Trends and Challenges
Edited by
Suresh Sundaresan
Professor of Economics and Finance, Columbia Business
School, USA
Edward Elgar
Cheltenham, UK • Northampton, MA, USA
© Suresh Sundaresan 2008
All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system or transmitted in any form or by any means, electronic,
mechanical or photocopying, recording, or otherwise without the prior
permission of the publisher.
Published by
Edward Elgar Publishing Limited
The Lypiatts
15 Lansdown Road
Cheltenham
Glos GL50 2JA
UK
Edward Elgar Publishing, Inc.

William Pratt House
9 Dewey Court
Northampton
Massachusetts 01060
USA
A catalogue record for this book

is available from the British Library
Library of Congress Control Number: 2008935924
ISBN 978 1 84720 920 7
Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall
Contents
List of contributors vi
Preface xi
1 The changing landscape of microfinance 1

Suresh Sundaresan
2 The role of international capital markets in microfinance 25
Brad Swanson
3 Securitization and micro-credit backed securities (MCBS) 46
Ray Rahman and Saif Shah Mohammed
4 Cell phones for delivering micro-loans 71
Anand Shrivastav
5 How should governments regulate microfinance? 85
Richard Rosenberg
6 Gender empowerment in microfinance 108
Beatriz Armendáriz and Nigel Roome
Index 123
v
Contributors
Beatriz Armendáriz is a Lecturer in Economics at Harvard University, and
a Senior Lecturer at University College London. She is also a research
affiliate at the David Rockefeller Center for Latin American Studies at
Harvard University. She has taught at the London School of Economics
and has worked as a Visiting Associate Professor at MIT, and as a Visiting
Professor at the Toulouse School of Economics. Her research focuses on
economic development and finance. Having published numerous articles on
microfinance, notably with Christian Gollier and Jonathan Morduch, she
recently co-authored The Economics of Microfinance, a book published by
MIT Press in 2005.
Her current research includes fieldwork on microfinance and gender
empowerment, with researchers from the Innovations for Poverty Action
(Yale and Harvard), and the Financial Access Initiative (Harvard, Yale,
NYU). She is also co-editing a handbook on microfinance, and a book, The
Contemporary Latin American Economy, also for MIT Press. Lecturer
Armendáriz grew up in southern Mexico where she founded AlSol and
Grameen Trust, Chiapas, the first Grameen-style microfinance organiza-
tions in the region.
Saif Shah Mohammed is co-founder of MR Analytics. In 2005–06, he
worked extensively on the BRAC micro-credit securitization, having moved
to Bangladesh to complete and implement the transaction. Prior to joining
MF Analytics, he worked as an analyst at Cornerstone Research, assisting
industry and faculty experts in developing economic and financial analyses
in litigation contexts. Saif graduated from Harvard College in 2002 with a
B.A. magna cum laude in economics. He currently attends the Columbia
University School of Law.
Ray Rahman is the founder and CEO of MF Analytics and one of the chief
architects of the BRAC micro-credit securitization. Prior to starting MR
Analytics, Ray was at Lehman Brothers, working on the cutting edge of
asset securitization in the fledgling commericial mortgage-backed secu-
rity (CMBS) industry in the 1990s. He has also working in private
equity, restructuring multinational companies with sales of over $1 billion.
Rahman is a cofounder of Potenco, a green energy company that is creating
the world’s most efficient hand-held power generator, targeted at the needs
vi
Contributors vii
of developing countries. The generator was featured in Wired magazine in

the summer of 2007. Rahman is also founder of Fast Forward India, a non-
profit organization that links graduate students from MIT, Stanford and
IIT with local mentors and NGOs, to identify and solve key problems that
directly affect the life of disadvantaged populations in India. FFI helps
teams pilot test proposed solutions on the field with local NGOs, obtain
funding to grow and widen the scale and scope of impact. Rahman gradu-
ated with a BS degree in mechanical engineering (high distinction) and a
BA in economics, (cum laude) from the University of Rochester. He has an
MBA from MIT Sloan School of Management.
Nigel Roome holds the Daniel Janssen Chair in Corporate Social
Responsibility at Solvay Business School, Free University of Brussels, and
the University Chair in Corporate Global Responsibility and Governance
at TiasNimbas Business School, Tilburg University, Netherlands. He previ-
ously held chairs in The Netherlands and Canada, and academic positions
in Britain. He is widely published on topics that relate business strategies,
innovation, and technology to issues of corporate responsibility, sustainable
development, and global change. His books include Management Education
for Sustainability and Corporate Environmental Management (1994);
Sustainability Strategies for Industry (1998); and The Ecology of Information
and Communications Technologies (2002).
Professor Roome’s career has involved innovations in both education
and research. His achievements as a European Faculty Pioneer over 20
years were acknowledged by The Aspen Institute in 2006. Roome is cur-
rently Chair of the Academic Board of the European Academy of Business
in Society and is a member of the academy’s management board. In addit-
ion to these responsibilities Roome has served as an environment coun-
cilor, an advisor to the president of Ontario Hydro and the Office of the
Canadian Commissioner for Environment and Sustainable Development,
a member of ABB’s global stakeholder advisory group, an expert to the
European Commission’s “Futures” project, and Chair of the European
Commission’s expert group on competitive and sustainable production
systems to the year 2020. He has been invited to contribute his views on
corporate responsibility and sustainability to meetings of the European
Union, under successive presidencies of Sweden, The Netherlands,
Finland, and Germany.
Richard Rosenberg holds a law degree from Harvard University. He has
worked in international development since 1984 with USAID and the
World Bank, concentrating on development finance, especially micro-
finance. He has authored or co-authored two dozen publications on
microfinance, including several on issues associated with regulation and
viii Contributors
supervision of microfinance, and teaches regularly on that topic at the

Boulder Institute of Microfinance Training.
Anand Shrivastav is the chief software architect and founder of Suvidha. He
has over 25 years of experience in marketing and distribution of consumer,
nutrition, and health care products besides telecom and banking services.
He has set up projects, implemented marketing strategies, and managed
large distribution channels. Shrivastav currently serves on the boards of the
following organizations: Suvidha Starnet (mobile transaction services), as
Chairman; and Genesis Biogen (stem cell technologies), as Chairman.
In the past Shrivastav has served Intercorp Biotech (biotechnology,
nutrition) as managing director, Hiperworld (information technology/
software) as managing director, The Home Store India (retail chain)
as director, Intercorp Industries (hybrid seeds) as director, IEPCL
(telecom, chemicals) as executive director, as marketing consultant for
public sector units Kerala State Drugs & Pharmaceuticals and Kerala
Soaps & Oils (consumer products), and as resident manager with RG
Soft Drinks. Shrivastav attended Harvard University Graduate School of
Engineering Design.
Suresh Sundaresan is the Chase Manhattan Bank Professor of Economics
and Finance at Columbia University. He is currently the Chair of the
Finance subdivision. He has published in the areas of Treasury auctions,
bidding, default risk, habit formation, term structure of interest rates, asset
pricing, pension asset allocation, swaps, options, forwards, futures, fixed-
income securities markets, and risk management. His research papers have
appeared in major journals such as the Journal of Finance, Review of
Financial Studies, Journal of Business, Journal of Financial and Quantitative
Analysis, European Economic Review, Journal of Banking and Finance, and
Journal of Political Economy, among others. He has also contributed arti-
cles in the Financial Times and to World Bank conferences. He is an
associate editor of Journal of Finance and Review of Derivatives Research.
His current research focus is on default risk and how it affects asset pricing
and sovereign debt securities.
Sundaresan has consulted for Morgan Stanley Asset Management and
Ernst and Young. His consulting work focuses on term structure models,
swap pricing models, credit risk models, valuation, and risk management.
He is the author of the book Fixed-income Markets and their Derivatives
(1996). He has served on the Treasury Bond Markets Advisory Committee.
He has served as a resident scholar at the Federal Reserve Bank of New
York. He has worked with the Center for Microfinance Research (CMFR)
on issues relating to interest rates on micro-loans, the effects of repeated
borrowings, and contract design.
Contributors ix
Brad Swanson is a partner in Developing World Markets, a socially respon-

sible investment bank and fund manager, which he joined in 2003. As a
banker and a diplomat, he has worked in emerging markets for 25 years and
has done business in more than 50 countries. He began his career in the US
State Department and was posted to West Africa. He then worked for
Donaldson, Lufkin & Jenrette Securities Corp. in New York, Bankers Trust
and Banque Nationale de Paris in London, and Global Environment Fund
in Washington, DC. During spring and summer of 2004, he was on assign-
ment with the US Department of Defense in Iraq running private-sector
finance programs for the occupation government. He is a member of the
Investment Committee of Partners for the Common Good, a non-profit
community development finance organization, and has served on the
boards of investment banks in Poland and the Ivory Coast. He has a BA
from Princeton University and an MBA from Columbia University.
Preface
This book—Microfinance: Emerging Trends and Challenges—brings
together recent practical innovations and policy questions in the field of
microfinance, and is largely based on the contributions made by different
scholars and practitioners to the conference hosted by the Social Enterprise
Program of Columbia Business School on 20 April 2007 at Columbia
University. This conference, entitled Credit Markets for the Poor: Focus on
Micro-Finance, examined recent developments in the field, research
findings, and the challenges that lie ahead. The contributions in this book,
after the introductory chapter, focus on (1) integration of capital markets
with microfinance, (2) securitization and micro-credit backed securities, (3)
technological innovations such as the delivery of banking services to the
“unbanked sector” using mobile phone technology, (4) the regulatory chal-
lenges and opportunities as the landscape of microfinance undergoes a sea
change, and (5) the consequences of gender empowerment, where women,
especially among the poorest, predominate micro-loan borrowings.
xi
1. The changing landscape of
microfinance
Suresh Sundaresan
INTRODUCTION
The ability of households to save, access capital, and manage risk exposures
of various kinds, such as life, property, and health through insurance is a
prerequisite for their economic and social development. Access to basic
financial services (such as credit, savings, and insurance) is most likely to
develop the entrepreneurial skills and opportunities among those poor who
are currently outside the perimeter of such financial markets and services.
Furthermore, over time, such access will promote better risk management
capabilities and promote the economic aspirations of the poor.
The World Bank uses two reference benchmark levels of consumption/
income to measure poverty: a consumption level of (US) $1.08 per day and
a consumption level of $2.15 per day.1 These levels are measured in 1993
purchasing parity terms. As of 2001, the World Bank estimated 1.1 billion
people had consumption levels below $1 a day, and 2.7 billion lived on less
than $2 a day. While these figures are very stark, it is also true that the pro-
portion of people living under $1 a day has fallen from 28 percent in 1990
to 21 percent in 2001. This progress not withstanding, it remains clear that
poverty alleviation should be a major priority, especially for countries where
a great proportion of people live under $1–2 a day.
It is difficult to visualize how countries such as Brazil, China, and India
could truly emerge as developing economies until their numerous poor citi-
zens find easy access to essential financial services, which is critical to them
in climbing out of poverty. Microfinance, which has emerged and evolved
over the past 35 years, is one such mechanism that has attempted to deliver
the core financial services to the poor. This mechanism along with some
recent developments in this field is the focus of this book.
It should be made clear that while the provision of low-cost access to
financial services is an important ingredient in alleviating poverty, there
are a number of other basic services that are unavailable to the poor. These
constraints include (1) absence of primary education, (2) absence of
1
2 Microfinance
primary health care, and (3) relatively primitive technologies used by the
households. These constraints need to be tackled in parallel as the momen-
tum to deliver financial services gathers speed. There are significant com-
plementarities between access to financial services and the ability of the
poor to access education, health care, and better technologies.
FORMAL AND INFORMAL MARKETS
Capital markets and financial institutions help provide mechanisms for

thrifty households to save and provide access to households and institu-
tions seeking capital for funding their consumption and investment plans.
In addition, they offer financial services such as life insurance, property
insurance, and health insurance. The successful development of economies
in Europe and the United States is in no small measure due to the services
provided by such institutions and markets, which allocate savings from one
part of the economy to finance the capital requirements in other parts.
Organized stock exchanges, bond markets, dealer markets, commercial
banks, insurance companies and other institutions, constitute a vital part
of financial architecture of most economies, whereby households and
institutions are able to allocate their savings and have access to capital.
These markets and institutions are known as formal markets for financial
services. Formal credit markets consist of institutions such as commercial
banks, credit unions, rural banks, and other financial institutions, which
are subject to public and private oversight and regulation. They are gov-
erned by bankruptcy codes, investor protection laws, and disclosure
requirements. In many developing economies, however, a significant
percentage of the population is simply unable to access these organ-
ized financial institutions and markets due to the fact that they are
extremely poor, ill-educated, and have no sustained opportunities for
gainful employment. From the perspective of the lenders in the formal
credit markets, delivery of small loans or provision of insurance to cover
very small individual exposures, or accepting tiny savings deposits is
simply not an economically viable proposition. This is due in no small
measure to the transactions costs associated with extending such tiny
scale of services to the poor, who are numerous.
Yet another important reason for lack of access of markets to the poor
is the fact that lenders and service providers in the formal markets have very
little information about the potential consumers of their services. As a
result, such poor households are unable to tap meaningfully into the formal
(organized) markets for financial services or capital markets for financing
their consumption and investment needs.
The changing landscape of microfinance 3
The properties of financial markets where there are significant differ-

ences between the information possessed by the lender and the information
possessed by the borrower have been well documented in a number of
influential papers.2 In particular, the following insights emerge from this
strand of research: first, with asymmetric information, there will be credit
rationing, leading to households being shut out of credit markets. Second,
contracts will have to be designed so that the lender (who has limited infor-
mation about the borrower’s ability or willingness to pay), can assure him
or herself a fair rate of return on loans. In fact, microfinance is an area
where the concept of “group lending” or “self-help groups” (SHGs) is
widely practiced: loans are extended to a group of borrowers who are
jointly liable for each of the loans extended to the members of the group.
Even the formation of the group itself is a matter of considerable import-
ance: since the borrowers have better information about each other, it is
recognized that the borrowers should be allowed to form the groups them-
selves. The presence of joint liability will then imply that “high-risk” bor-
rowers will not be accepted in the group, and the pool of “low-risk” groups
will then also be conservative in the choice of projects.
The loan contract itself may be structured to reflect ground realities: loan
contracts may tend to have short maturity in order to set strong incentives
for the group to service the debt payments, where the income stream is
steady. Payments of initial rounds of loans may qualify borrowers to obtain
a loan of higher size in the subsequent rounds: this provides an additional
incentive to borrowers to acquire good credit history. Interest payments may
be collected on a weekly basis, thereby assuring the presence of loan officers
at site levels at that time. In seasonal projects, loan payments may be indexed
to reflect the revenue patterns. Loans may be bundled with insurance pro-
grams to mitigate aggregate risk exposure such as monsoon or droughts.
In the past and to a significant extent even now, poor households have
tended to rely extensively on informal markets for their capital needs and
other financial services such as insurance or savings. Indeed, these other
financial services have only recently become available to poor households
through microfinance. Informal credit markets typically operate outside the
perimeter of regulators and are often not subject to monitoring and super-
vision by governments or agencies of governments. The rights and respon-
sibilities of lenders and borrowers in such markets do not come under a
formal bankruptcy code.
Players living in close proximity to such markets often have detailed
knowledge of each other. Examples of such informal markets and
participants might include the following: (1) family members, (2) friends,
(3) trade credit from local shops, (4) professional moneylenders in the
region, (5) pawnbrokers, (6) local landowners, and so on. These markets
4 Microfinance
have existed for a long time, and indeed have preceded the development of
formal banking systems and capital markets. Because they provide a very
useful function in poorer sectors of world economies in keeping the savings
flow from lenders to borrowers who are unable to access formal credit
markets such markets will continue to exist. In this context, it is worth
pointing out that even in a well-developed economy such as the United
States, pawnbrokers and payday lending institutions exist and serve a clien-
tele of borrowers.3
Without necessarily detracting from the useful functions provided by
such informal credit markets, it is also important to examine more carefully
the interest rates that prevail in them. Roughly, interest rates in such infor-
mal markets provide an upper bound on the interest rates that the borrow-
ers would be willing to pay in microfinance markets. While detailed and
reliable micro-level data on such markets are usually not available, several
studies have documented that the effective borrowing costs in such infor-
mal credit markets are rather high. The effective annualized interest rates in
payday lending runs well into three digits, often in excess of 200 percent on
an annualized basis! For example, interest rates estimated in pawnbroking
and local moneylenders run into triple digits, on an annualized basis.
In the United States, interest rates charged by pawnbrokers ranged from
36 percent (in New Jersey and Pennsylvania) to 240 percent (in Oklahoma)
during 1987–88.4 In addition, the supply of capital from informal credit
markets tends to be rather limited. These observations suggest that a preva-
lence of high interest rates in informal credit markets where the private
sector individuals possessing asymmetric information play the role of
lender of last resort to liquidity-constrained households.
EVOLUTION OF MICROFINANCE
Informal credit markets described in the previous section have preceded

microfinance. Institutions such as credit unions and specialized lending pro-
grams targeted to agricultural sectors have also been in existence since the
early 1900s.5 The seeds for microfinance in its current form were planted
during the period 1950–80, when small loans were extended to poor bor-
rowers who could not post meaningful collateral. Major organizations, which
pioneered this initiative, were ACCION International in Latin America,
SEWA Bank in India and the Grameen Bank founded by Muhammad
Yunus in Bangladesh. These initiatives demonstrated for the first time that
poor borrowers, especially women, were not only willing to take on small-
scale projects funded by loans, but were also capable of chalking up excel-
lent payment records.
Over the period 1980–90s many microfinance institutions began to

develop and found sustaining models of lending to the poor: non-
governmental organizations (NGOs), non-bank financial institutions
(NBFI), rural banks of nationalized banks, and village banks began to
develop.6 In the early stages of evolution, microfinance was largely
restricted to loans and was funded by either governments or aid agencies
and was thus based on “soft capital.”
Since the 1990s, microfinance has branched out both in terms of the
range of financial and economic services extended, as well as in terms of
how capital is raised. Banks began to access this market in a more
significant way than ever before. Financial services ranging from savings
deposits, loans, insurance to cover life, health, crop, and properties are cur-
rently offered. Many microfinance institutions access capital markets either
by issuing equity or debt capital in order to raise capital. There are others
who have been able to securitize their loans and thus attract capital by
issuing micro-credit backed securities.
Technological innovations have also paced the evolution of microfinance:
widespread availability of mobile phones, access to community-level kiosks
of computer terminals with access to the Internet, biometric technology to
obtain loan approval and credit history, and correspondent banking have
dramatically changed the landscape of microfinance.
The year 2005 was declared as the “Year of Microfinance,” and a number
of private sector enterprises and foundations have now dedicated pools of
capital for exclusive investments in the area of microfinance: in November
2005, Pierre Omidyar (founder of eBay) announced a $100 million
microfinance fund in partnership with Tufts University for exclusive invest-
ments in the microfinance initiatives. TIAA-CREF created in September
2006 a $100 million Global Microfinance Investment Program (GMIP) to
invest in selected microfinance institutions worldwide. TIAA-CREF made
a $43 million private equity investment in ProCredit Holding AG, a
microfinance company.7 These are high profile, large-scale investment ini-
tiatives. Given the current size and status of microfinance institutions it will
be interesting to see how these initiatives pan out.
As Swanson and Rahman and Mohammed point out respectively in
Chapters 2 and 3, concepts such as securitization are rapidly integrating
microfinance with the capital markets of the developed world, thereby
altering the landscape of microfinance. As we will demonstrate later in this
chapter, traditional institutions such as NGOs still remain an important
force, especially for the poorest of micro-borrowers. In the upper tier of the
microfinance, capital markets are actively helping microfinance institutions
to tap debt and equity capital. New market institutions have developed,
which are likely to improve the transparency and potentially reduce the cost
6 Microfinance
of accessing financial services: for example, a recent study has reported that
the microfinance institutions (MFIs) in countries with credit bureaus tend
to have a 5 percent lower operating expense ratio than the ones in countries
without credit bureaus.8 Credit rating agencies have developed in countries
that rate institutions in microfinance. Two rating agencies, MicroRate and
M-CRIL are well established in the field.
QUESTIONS ADDRESSED IN THIS BOOK AND A

WORD ABOUT DATA
What are the costs of accessing credit in microfinance? Does it depend on

the nature of contracting? Does the nature of organization of the
microfinance institution affect the interest rates, and the characteristics of
the borrowers? How have the capital markets and financial intermediaries
influenced the character of microfinance? What are the risks and returns to
institutions in microfinance? What are the implications of technological
innovations such as biometrics, the Internet, mobile phones, and so on, on
microfinance practice? Is there a strong gender bias? What are the positive
and adverse consequences of such a bias?
These questions form the intellectual basis for this book. To motivate
these issues, I would like to examine some of the studies that have used MIX
data, in order to shed some light on the issues studied here. Before delving
into these questions, it is useful to note the following properties of MIX
data. MIX data is based on voluntary reporting by 704 institutions to a
detailed survey conducted by MIX. As of 2006, the survey covered more
than 52 million borrowers with $23 billion in loans. The survey also covered
56 million depositors with over $32 million in deposits. Clearly, these
figures are a downward-biased estimate of the true size of the microfinance
market. It is also conceivable that the participating institutions are qualita-
tively different from the non-participating institutions, which are likely to
be smaller and more donor-financed. With these caveats, let us examine the
questions in turn in the following sub-sections.
Cost of Access to Credit
Lacking micro-level data, it is difficult to estimate the costs of obtaining

financial services to the borrower in the microfinance area with great pre-
cision. Still, we have some anecdotal evidence about the levels of interest
rates that micro-loan borrowers encounter. In a recent paper, CGAP
(Consultative Group to Assist the Poor) reports that the interest yield for
Compartamos (a lender in this market) stood at 86.3 percent!9 CGAP
estimates that the cost to the micro-loan borrower is about 100 percent
when taxes are taken into consideration. This data corresponded to the
2005 period when the median interest rates charged by village banking
MFIs stood at 47.2 percent, and the interest rates charged by low-end MFIs
stood at 35.4 percent.
Gonzalez (2007) provides an analysis of the MIX database, which
enables us to shed some light on this issue at a more aggregate level,10 and
using this data, we can estimate the interest costs faced by the borrower.
Our first estimate is the yield on gross loan portfolio. This measure takes
the ratio of adjusted financial revenue from the loan portfolio to the
adjusted average gross loan portfolio. Financial revenue includes the
revenue from the loan portfolio and other assets plus revenue from other
financial services. Such services may include insurance, passbooks, smart
cards, and so on. For our purposes these expenses are relevant to the mea-
surement of the costs of financial services. Gross loan portfolio excludes
write-offs. In Table 1.1, I provide an estimate of the costs for different
lending institutions for the period 2003–05.
Note that the costs range from a low of about 20 percent to a high of over
42 percent. The estimated costs are the highest for NGOs, which service the
poorest of the borrowers, and lowest for credit unions. Transactions costs
associated with delivery of loans, monitoring, collection, and the risk of
default are the primary reasons for such high interest rates. Gonzalez (2007)
in his analysis also provides another way to look at this measure: across
different target group of borrowers,11 as shown in Table 1.2.
Table 1.2 confirms our finding that the low-end borrowers with a loan
balance of less than $150 are the ones who face the highest cost of obtain-
ing financial services. Their costs range from 35 percent to 38 percent. This
raises the question of whether at these rates it is reasonable to think that
microfinance can play an effective role in alleviating poverty? There are
several reasons to harbor such a hope. First, it is safe to assume that the
Table 1.1 Costs of obtaining financial services—variations across

institutions (in %)
Institution 2003 2004 2005

Banks 33.7 36.9 28.3
Credit unions 22.0 20.2 20.4
NBFI 32.6 32.1 29.3
NGOs 42.5 40.2 38.6
Rural banks 27.3 26.8 23.2
Source: Gonzalez (2007).

8 Microfinance
Table 1.2 Costs of obtaining financial services—variations across target

groups (in %)
Target Group 2003 2004 2005

Low end 37.8 37.8 35.4
Broad 34.5 33.7 31.1
High end 24.1 24.1 22.7
Small business 24.8 22.7 22.6
alternative sources of credit and other financial services are even more
expensive.12 Since the participation in microfinance is purely voluntary, it
is reasonable to conclude that for the participants this avenue must be
cost-effective.
Second, the costs appear to decline for the broad and higher-end
borrowers. Presumably these are seasoned borrowers, who by repeated
borrowing have established a good credit reputation and honed their
entrepreneurial skills. This has in turn dramatically declined the costs of
access.
Finally, there are reasons to think that the average costs of participating
in this market is expected to go down significantly due to the advent of tech-
nology such as biometric screening, smart cards, and the delivery of loans
by mobile phones. This said, it is clear that the borrowing costs must come
down significantly in order for microfinance to be a credible and sustaining
avenue for poverty alleviation.
Contracting and cost of access

As noted earlier, lenders utilize different contracting methods to extend
loans. Some borrowers get “individual loans” and others are part of a “sol-
idarity group” where the group is jointly liable for the loans taken by the
members. Table 1.3 reports the cost of accessing loans during the period
2003–05 across different contracting methods.
Table 1.3 shows that the interest rates are highest for village banking
groups which consist of poorest borrowers formed into solidarity groups.
We note that “individual borrowers” face borrowing costs ranging from 28
percent to 30 percent, among the lowest.
Capital structure and default risk

The cost of extending financial services such as credit and insurance reflects
the costs of accessing capital for different lending institutions and the risks
Table 1.3 Costs of obtaining financial services—variations across

contracting (in %)
Contracting Method 2003 2004 2005

Individual 30.1 30.7 28.5
Individual/Solidaritya 33.8 33.4 30.7
Solidarity 35.3 40.4 35.9
Village banking 52.0 49.6 47.2
Note: a. Group lending.
of the consumers of financial services. We now turn to these issues. In the

analysis of trend lines, Gonzalez (2007—see note 10) reports that during
the period 2003–05, the debt/equity ratios of banks varied from 3.6 to 5.6.
The corresponding range for credit unions was 3.4 to 4.4. For the rural
banks, the debt/equity ratios ranged from 4.6 to 5.2. In contrast, non-bank
financial institutions and NGOs had much smaller debt/equity ratios. The
low debt/equity ratios of NGOs (ranging from 0.9 to 1.6) may reflect the
pervasive use of donor capital by NGOs. High debt/equity ratios of banks,
credit unions, and rural banks may reflect their inability to get sufficient
equity capital and the relative ease of access to debt capital. In the next sub-
section, we provide some evidence concerning the ability of microfinance
institutions to tap into debt and equity capital.
Role of Capital Markets in Financing and Investment in Microfinance
Chapters 2 and 3 explore the role of capital markets in the field of

microfinance. Microfinance began largely as a philanthropic effort or a
quasi-philanthropic effort. Government-mandated programs such as rural
banks, and branch expansion by nationalized banks into rural areas, are
examples of such effort. Non-governmental organizations (NGOs), whose
supply of capital is largely donor-based, dominated the scene. In fact, an
analysis of the MIX database shows that there are over 400 NGOs and over
100 rural banks that extend financial services as of 2006. Indeed, this evi-
dence demonstrates the very key roles that these groups continue to play in
microfinance. But we have seen a very active interplay between capital
markets and microfinance: increasingly, capital markets are being used to
source capital for providing microfinance services. Indeed, we tend to see a
tiered market: some of the top-tier microfinance institutions are able to
access capital markets for fairly large chunks of capital.
10 Microfinance
There are other microfinance institutions that appear to rely on domes-

tic and local markets for their funding needs. In Chapter 2, Swanson
explores this development and the challenges and opportunities that lie
ahead. In Chapter 3, Rahman and Mohammed explain in detail the first
micro-credit backed securitization, which tapped capital from different
financial institutions to the microfinance sector in Bangladesh.13 These
developments show that microfinance is starting to establish itself as a
major sector in the eyes of big financial institutions. In due course, this will
mean that the assets backed by micro-loans may potentially establish them-
selves as a separate asset class.
In addition to the securitization deals that are described by Swanson and
Rahman and Mohammed, we have also seen other major developments in
the interaction between capital markets and microfinance institutions. In
2006, shares of Equity Bank in Kenya were listed on the Nairobi Stock
Exchange. This is reputed to be the first microfinance IPO in Africa, sig-
naling the ability of some institutions in microfinance to access equity (risk)
capital.14 On 20 April 2007, Banco Compartamos, a microfinance institu-
tion launched in 1990, made the first initial public offering (IPO).15 This
implies the flow of equity capital into microfinance from the capital
markets. Major international financial institutions (IFIs) such as EBRD,
IFC, and KfW raise significant amounts of capital and invest in the field of
microfinance. These IFIs are private-sector arms of public financial insti-
tutions. A recent report suggests that IFIs have a portfolio of $2.4 billion
in microfinance, with the average investment size at $4 million.
Microfinance investment vehicles (MIVs) have been created to channel
private sector funds into microfinance. The MIVs have a portfolio of $2
billion, with the average investment size of $1 million. This is a topic ana-
lyzed by Swanson in Chapter 2.
Securitization techniques are used to create debt and equity financing
possibilities for microfinance institutions. For example, Citigroup recently
created $165 million of collateralized debt obligations (CDOs), which will
be initially backed by 30 MFI loan portfolios in 13 countries. Private
investors are expected to invest $69 million of senior tranches rated AA by
Fitch. IFIs and Citigroup will invest in the subordinated tranches and
equity.16 Swanson analyzes the CDO structures and the benefits that accrue
from such innovative financing methods. Rahman and Mohammed in
Chapter 3 analyze the BRAC securitization deal in sufficient detail so that
the reader can appreciate the benefits of such innovative ways to tap into
capital markets.
A migration has also started to occur, whereby NGOs are starting to shed
their “not-for-profit” status and become regulated providers of financial
services.17 Stephens (2007—Note 17) reports the growth of deposit-taking
services and increased commercialization, with the median commercial

funding of loan portfolios increasing from a level of 40 percent in 2003 to
60 percent in 2005. This also presents interesting regulatory challenges on
issues such as governance, investor protection, disclosure requirements, and
so on.18
Technology and Microfinance—Opportunities and Challenges
In an earlier sub-section, we noted that the borrowing rates in microfinance

markets can be rather high. In addition, we noted that the delivery of
financial services such as tiny loans and insurance payments, and accepting
tiny savings deposits require that the service providers overcome (1)
significant transactions costs, (2) adverse selection problems, and (3) moral
hazard issues. In fact, a skeptic could reasonably ask how microfinance could
be the mechanism for alleviation of poverty, if the interest rates are so high,
and if scalability is simply unattainable due to the factors outlined above.
Technological developments over the last decade hold much hope in
overcoming these obstacles. I will review some of these developments in
order to set the stage for Chapter 4, in which Shrivastav explains how
mobile phone technology is used in India to potentially deliver loans that
are as low as $20.
Electronic matching of borrowers and lenders: search and delivery costs

One of the factors that contributes to the relatively high interest rates on
loans is the cost associated with search and delivery of loans, and the sub-
sequent efforts that are needed to manage the default risk of the loan port-
folio. The Internet has opened up new vistas for matching borrowers and
lenders electronically, and the developments that we have seen in the
last decade have the potential to lower the costs of very small loans. In
Table 1.4 we highlight some of the recent innovations that have occurred in
the electronic matching of borrowers and lenders.19 There are firms such as
Prosper.com in the United States, and Zopa.com in the United Kingdom,
which permit peer-to-peer micro-lending and essentially avoiding any inter-
mediaries.20 These options are alternatives for the small borrower to piling
up credit-card-type debt. On these websites potential lenders are able to
view and evaluate loan requests, ranging from about $1000 upwards, and
bid on them. Most of the loans are of short duration, not exceeding three
years. Legally binding contracts are entered into. Information about credit
scores is used in contracting. Monthly payment schedules are enforced.
Unsurprisingly, interest rates on these websites are well above what one
sees in the “formal credit markets,” ranging anywhere around 15 percent to
25 percent or more.
12 Microfinance
Table 1.4 Matching borrowers and lenders—some innovations
Organization Services Provided Some Attributes

KIVA Kiva enters into a Lenders are able to select
partnership with existing their borrowers and their
microfinance institutions loans are electronically
worldwide to link lenders transmitted to the local
and borrowers MFIs
Zopa Zopa partners with credit Zopa is able to offer the
unions thus, it offers attractive features of credit
deposits as investments, and unions through the Internet
lending opportunities to its investors, lenders, and
borrowers
Prosper market Provides an open market Has several specialized
place place to match lenders and lending programs targeted
borrowers. Diversification at different communities
of loans is encouraged
Lending Club Borrowers complete a Accessible to members in
personal loan request and the network
are screened based on a
minimum FICO score, and
so on. Interest rate is fixed for
three years
Source: Author’s compilation.
Table 1.4 shows a group of organizations that have exploited the Internet
to reduce dramatically the time it takes to match borrowers and lenders,
and to arrange the flow of loans between them as well as to promote social
lending at a profit.21 This trend may well expand more in to the field of
microfinance, although there are barriers such as illiteracy, and inability to
access community networks of computers to access the Internet. These bar-
riers may well yield due to other technological advances, which we sketch
next.
Correspondent banking and biometric authorization of credit22

A development that has accelerated the access of financial services to the
poor is the concept of “correspondent” banking or branchless banking.
This has been especially successful in Brazil, and the concept is bound to
have a major effect in other regions.23
Under this concept, local post offices and shops are equipped with
barcode-reading point-of-sale (POS) terminals. These local entities then act
as agents for banks. It is estimated that currently there are nearly 100 000
such correspondent entities in Brazil alone. CGAP estimates that nearly 13
million customers have been brought into the fold of the banking system
through these correspondent networks.
It is well recognized that the literacy levels of microfinance borrowers are
rather low. This presents unique challenges in reducing the costs of lending
to them. One technological development that has entered the field of
microfinance is the application of fingerprints for the purpose of ident-
ifying and validating financial transactions, through ATM networks.
Biometrics and smart cards are already in use in microfinance in India and
Indonesia. Biometric teller machines (BTMs) reduce the administrative
costs of extending small loans in communities where the literacy levels
are low. The smart card contains the credit history of the borrower. Banks
such as ICICI in India and Danamon in Indonesia are employing such
technologies.
Village Internet centers and eChoupals

One of the developments that has occurred in the delivery of services to
poor citizens in remote corners of the world is the proliferation of com-
munity Internet centers, which deliver both the infrastructure and infor-
mation that may be utilized by small agricultural producers to get the best
possible price and thus eliminate intermediaries.24 Some see this develop-
ment as using the technology to empower as well as significantly increase
the range of services to the poorer sections of society. One such innovation
is the concept of eChoupals, pioneered by India Tobacco Company (ITC),
which deliver price information and an infrastructure for buying the agri-
cultural production from small farmers. eChoupals are village Internet
kiosks run by local entrepreneurs who provide pricing information for
different delivery dates to local farmers and enable the farmers to sell their
produce directly to ITC, bypassing the intermediaries. Farmers benefit
from a known price schedule, and ITC benefits from the elimination of
commissions and transactions costs that intermediaries would have
charged.
Mobile phone delivery of financial services25

It is estimated that there are over 2 billion mobile phone users who not have
bank accounts.26 In recent times, many companies have been able to utilize
mobile phone technology to provide transfer of cash, make loans, and
extend basic financial services. This promises to bridge the “unbanked”
sector with the market for financial services at a speed that could not have
been imagined a decade ago. In South Africa, it is reported that banks have
been able to link a debit card and a bank account to a cell phone. This
14 Microfinance
enables the owners of cell phones to make a deposit at a bank or any post
office, and the deposits are then credited to an account and confirmed via
text message.
USAID has pioneered mobile-phone-based access to financial services
for the poor in the Philippines. Microfinance customers make loan pay-
ments by a text messaging system, dramatically lowering the transactions
costs and eliminating intermediaries in the process. A concept known as G-
cash has been implemented in the Philippines, which effectively allows the
users of cell phones to send and receive cash via text messages. As of March
2006, over 1.3 million customers are using the G-cash system, which
handles $100 million a day. This avenue promises to cut the transactions
costs, time, and effort for both borrowers and lenders. Since the growth of
cell phone customer base has been exponential in the last decade, the poten-
tial for a steep drop in the cost of delivery of financial services through
mobile phones is promising.
This particular technological breakthrough is the topic explored by
Anand Shrivastav in Chapter 4. Shrivastav examines the growth of
mobile phone technology in India and its potential for delivery for financial
services. He also examines the different players in the market. He then
describes a technology that he has developed for the delivery of micro-
loans.
Regulating Microfinance
We have traced some of the major changes that have occurred in the land-
scape of microfinance. These changes present regulatory challenges, which
form the focus of Chapter 5 in which Rosenberg explores a framework for
regulating microfinance. One such challenge is the question of how to inte-
grate mobile-phone-based delivery of credit into the overall banking
system with appropriate safeguards. In addition, other challenges arise
from lender–borrower relationships. We have noted that the costs of access-
ing credit in informal credit markets in general and in microfinance in par-
ticular can be in the range of 20 percent to in excess of 50 percent. Our
analysis also showed that the poorest of the borrowers are often subject to
the highest of the interest rates, and they are often in group lending (self-
help groups—SHGs), obliging them to expend greater resources in peer
monitoring.
A question that naturally arises is whether there must be some oversight
on the interest rates charged by microfinance lending organizations. Since
the borrowers are often not well educated, another question is whether they
fully understand the effective interest rates that are being charged by the
lending institutions. Such effective interest rates will have to reflect a
number of factors such as: (1) frequency of compounding and payments,

(2) transactions costs charged, (3) costs of any mandated insurance poli-
cies, and (4) efforts associated with compliance—peer-monitoring efforts,
time spent with loan officers, and so on. Full disclosure of these factors may
not always take place.
A related issue is the social pressure that may be placed upon the bor-
rowers to enforce timely payments: a borrower in a group-lending contract
faces considerable pressure from the community to service the loan pay-
ments, as there is (1) joint liability, and (2) he or she lives in the community,
and default can qualitatively and adversely affect his or her social life.
Absence of credit bureaus and other institutions may also contribute to
a situation whereby the same borrower may end up obtaining multiple
loans well beyond his or her ability to service the loan payments. This may
result in defaults, which may adversely affect otherwise financially healthy
microfinance lenders.
Defaults may arise due to many reasons. A borrower may default because
he or she is unable to service the loan contract due to insufficient income.
Such poor realizations of income may be the result of factors outside the
control of the borrower (such as drought, monsoon, and so on) or owing
to poor incentives or insufficient effort. Defaults may also be strategic and
may be coordinated by the entire borrowing group. The risk of default is
managed by the lenders through various methods: (1) forming groups that
are less risky, (2) contracting with built-in incentives for peer monitoring,
(3) frequent (often weekly) payments where feasible, (4) extensive lender
monitoring, and so on. Some recent developments, related in the next sub-
section, point to some underlying problems that call for prudent oversight
of microfinance practice.
Defaults in 2006 on micro-loans in Andhra Pradesh27

The state government of Andhra Pradesh (in India) shut down 50
branches of two major MFIs in the Krishna District in March 2006. This
extreme action was precipitated when many micro-loan borrowers com-
plained to the state government about the “usurious interest rates” and
“forced loan recovery” practices. There was an allegation that ten borrow-
ers of MFIs in the Krishna District committed suicide because of their
inability to repay the loans taken from the MFIs. MFIs operating in
Andhra Pradesh reached an agreement with the state government on MFI
interest rates, product portfolio, inter-MFI competition, credit disburse-
ment, and loan recovery methodologies. As per the terms of the agree-
ment, MFIs have agreed to an interest rate ceiling of 15 percent. They have
agreed to desist from providing multiple credit to an existing borrower and
recovery of loans at a pace compatible with the borrower’s income level.
16 Microfinance
MFIs are also to remain strictly within the micro-credit domain, avoiding
micro-insurance products.28
This episode raises several important regulatory questions: should the
government set a ceiling on interest rates charged by microfinance lenders?
How often should such ceilings be reviewed and reset to reflect credit
market conditions? Should the government set standards and enforce such
standards on acceptable loan recovery practices? This episode is a water-
shed in focusing attention on the need to have appropriate institutions and
lending and loan recovery standards in place in order to promote private
capital flow into underserved communities.29
Savings and intermediation

Many microfinance institutions also accept savings deposits and provide
other services such as insurance. The direct and indirect costs associated
with the provision of these services should be transparent to the consumers
of these services.30 Perhaps more importantly, this raises the question of
deposit insurance. Should there be a deposit insurance program? What will
happen to the deposits if a microfinance institution were to default? To get
a sense of the magnitude of this problem, let us take a look at the question
of how pervasive deposits in microfinance are. Gonzalez (2007—see
note 8) reports that the ratio of deposits to loans ranges from 30 percent to
46 percent for individual loans. For solidarity groups and village banking,
deposits are not reported, and may be presumed to be negligible. Deposits
are a big part of banks, credit unions, and rural banks. Obviously, they are
not an issue for non-bank financial institutions and NGOs. How should the
deposits of the poor be protected?
Technological innovations
Mobile phone delivery of financial services, community Internet portals
that enable farmers to sell their crops and so avoid intermediaries, and other
such innovations also raise important regulatory questions. Should mobile
phone deliverers of loans be given a banking license? We noted that in cor-
respondent banking, third parties interact with customers to deliver
banking services. Some of the issues that arise from a regulatory standpoint
include the following:
1. Should third parties (correspondents, for example) be permitted to

execute banking transactions and interact directly with customers? In
this context, how should electronic money transfers and mobile phone
transfer of cash be treated?
2. After 9/11, the regulatory practices in organized credit/banking
markets have brought into play risk-based anti-money laundering
(AML) rules, as well as rules for combating the financing of terrorism

(CFT) adapted to the realities of remote transactions conducted
through agents.31 How should they be adapted in microfinance?
Finally, issues of consumer protection also have to be addressed.

Regulation is costly from the perspective of both formulation and enforce-
ment from the regulator’s point of view. If poorly framed, it can also be
costly to comply with from the standpoint of the lending institutions in
microfinance. Rosenberg addresses this very important issue in Chapter 5
and articulates how we must evaluate the challenges in this regard.
Gender Empowerment and its Characterization
One of the striking facts about the field of microfinance is that an over-
whelming number of the borrowers are women. In Chapter 6, Armendáriz
and Roome examine this strategy of targeting women or the so-called issue
of gender empowerment in microfinance and note the salient fact that
indeed, in the aggregate, seven out of ten microfinance clients are women.
This predominant empowerment in lending has been examined in the liter-
ature, and many beneficial effects that arise from such an empowerment have
been documented. More recent evidence, though anecdotal, points to some
potential dysfunctional consequences of such an empowerment. In their
chapter, Armendáriz and Roome observe that we have no reliable empirical
evidence to examine the results of this gender strategy on the extent and
quality of economic and social development, and argue for future research
to examine this issue in greater detail. They also explore the potential
benefits of women bringing their male partners into the fold of microfinance
on a voluntary basis. Such efforts may reduce domestic friction, reduce
default, and potentially increase overall welfare. These and other important
issues raised by Armendáriz and Roome warrant additional research.
Rather than summarizing the potentially important issues that
Armendáriz and Roome raise in Chapter 6, it is useful to characterize the
nature of gender empowerment in greater detail:
● Is this empowerment related to the nature of the lending institution?

● If so, what are the potential causal factors?
● Do we tend to see greater gender empowerment when the average
pool of borrowers is poorer?
● Is it the case that poorer women borrowers tend to gravitate to certain
types of lending institutions?
● Are there geographical variations in gender empowerment? What are
the cultural factors that might explain such geographical variations?
18 Microfinance
Table 1.5 Gender empowerment—percentage of women borrowers across

lending institutions and regions
Region Banks Cooperatives and Non-bank NGOs

Credit Unions Financial
Institutions
Africa 52 39 62 81
East Asia 61 N/A 86 98
Latin America 52 52 52 71
South Asia 97 99 65 100
Source: MIX data of 2006 and author’s calculations.
Evidence on gender empowerment

Establishing such stylized facts may enable us to further expand on the
questions that are analyzed by Armendáriz and Roome. To this end, we
analyze the MIX data of 2006 to shed some light on these questions.32
Table 1.5 documents the distribution of the fraction of women bor-
rowers in the sample, split across geographical regions and the types of
lending institutions.33 While this is admittedly a very aggregated picture,
which may potentially mask country-specific sociocultural dimensions, it
still provides an interesting breakdown. The data clearly show that the
gender empowerment is extremely strong in South Asia, and to a lesser
extent in East Asia. It is much less so in Africa and Latin America. Indeed,
for micro-loans extended by banks, cooperatives, and credit unions,
and non-bank financial institutions, we find that the borrowers are very
nearly evenly split across the gender in Latin America, and roughly so in
Africa.
Sorting the results by the type of lending institution leads to a very inter-
esting stylized fact: NGO borrowers are predominantly women across all
regions. We will see later, unsurprisingly, is that NGO borrowers are typi-
cally the poorest of micro-borrowers and this is where we find the greatest
of gender empowerment. Gonzalez (2007—see note 8) in his trend line
analysis has also documented a very similar picture for the period 2003–05,
which is summarized on a more aggregated level in Table 1.6.
Microfinance borrowers are targeted for credit either on an individual
(stand-alone) basis or as a part of a solidarity group (group-lending) basis.
The benefits of a solidarity group and the associated peer monitoring
effects have been addressed in the literature.34 In addition, there is the
concept of village banking, which is another self-help group (SHG)
consisting of very poor borrowers, with low per capita income. SHG-based
Table 1.6 Percentage of women borrowers across different institutions

2003–05
Year Banks Credit Unions NBFI NGO

2003 52.8 73.0 53.8 79.0
2004 50.0 56.5 60.8 82.1
2005 52.5 60.0 56.1 79.7
Table 1.7 Percentage of women borrowers across different contracting

arrangements 2003–05
Year Individual Individual/Solidarity Solidarity Village Banking

2003 47.9 67.9 82.0 90.3
2004 53.9 66.2 92.0 94.5
2005 51.8 62.0 100.0 90.2
lending has the beneficial impact of peer monitoring and assortative

matching, which leads to lower default risk and hence should result in lower
borrowing rates. What we see in the data as documented in this chapter,
however, is that the SHG borrowers are among the poorest, and this causes
the interest rates to actually go up.
Table 1.7 clearly shows the concentration of women borrowers in the
SHG category and much less so in individual loan programs.
Median loan size

Table 1.8 shows that the median loan size of the NGO borrowers is, by an
order of magnitude, smaller in each region when compared with other bor-
rowers. South Asian borrowers are all uniformly poor, with a median loan
balance ranging from $100–130 across different lending institutions. In
light of this it is interesting that the gender empowerment is so different for
non-bank financial institutions in South Asia.
The median borrower in Latin America has a loan balance, which is four
to ten times the median loan balance of borrowers in South Asia: this
appears to suggest that the gender empowerment declines with increases in
average loan balance, which may be an instrumental variable for seasoned
(proven) borrowers, lower default risk, and greater entrepreneurial skills.
These assertions remain to be tested.
20 Microfinance
Table 1.8 Median loan size in US$ across lending institutions and regions

Institutions
Africa 529 373 215 129
East Asia 562 N/A 235 90
Latin America 1445 1510 1039 437
South Asia 127 124 126 107
Source: MIX data (2006) and author’s analysis.
Median total assets

Table 1.9 is provided to give the reader a sense of the relative dominance
of different lending institutions in the microfinance space. A caveat is in
order in this context: only four banks reported from South Asia, and
Grameen Bank, which operates out of Bangladesh, reported assets worth
$819 830 340 and is by far the biggest player. But this is an exception and not
the rule, as the median total assets of banks in South Asia is only
$33 709 260. In general, there are fewer reporting banks in the market, but
they dominate the assets (loans, for the most part) in the market. In South
Asia, the median assets of banks are nearly five times the median assets of
NGOs. In Latin America the corresponding multiple is 46, and in East Asia,
the corresponding multiple is 550! These numbers point to the increasingly
dominant role played by banks in extending financial services to poor bor-
rowers. As we noted in the section on “Technology and Microfinance,” with
the advent of Internet and mobile phones, we will begin to see an increasing
presence of banks and new players in this market. This has far-reaching regu-
latory implications: the median loan sizes are far higher for banks, and the
gender empowerment is far lower for banks, with the exception of South
Asia.
The issue of gender empowerment in microfinance requires a careful
study at two levels: first, at a suitable level of aggregation (perhaps at a
country or a regional level) in order to better understand how it comes
about and what the evidence is from a development perspective. Several
interesting questions arise at such a broad level of aggregation: perhaps
certain types of lending institutions prefer women borrowers. Or, it could
be the case that women borrowers gravitate towards certain lending insti-
tutions. Are there differences at the level of countries? If so, what are the
determinants of such empowerment? For example, with the exception of
Bangladesh, in countries with a majority Islamic population, we tend to see
Table 1.9 Median total assets in US$ across lending institutions and
regions

Institutions
Africa 19 980 000 1 915 690 3 006 357 1 809 247
East Asia 742 857 915 N/A 6 237 324 1 350 415
Latin America 166 225 000 11 394 423 29 777 549 3 560 909
South Asia 33 709 260 1 143 689 8 516 418 6 098 412
Source: MIX data (2006) and author’s analysis.
fewer women borrowers in the MIX database, but NGOs seem to have suc-
ceeded in attracting women borrowers in these countries.35
At a second level, we need to address this issue at a very micro level: do
we see group-based lending and greater gender empowerment going hand
in hand? Is there a significant positive association between first-time bor-
rowers (who are likely to be among the poorest) and gender empowerment?
I find the issues studied in the chapter by Armendáriz and Roome to be
very important in many respects. First, we want to document clearly the
developmental and social benefits arising from targeting women in
microfinance to guide future efforts. Second, we need to better understand
why this empowerment does not appear to be present with some lending
institutions in some countries, at least in the context of the evidence pre-
sented from the MIX database. The potential benefits of voluntary intro-
duction of male partners by women into a microfinance program as a way
to alleviate the number of important issues highlighted by Armendáriz and
Roome (in Chapter 6) seems to be well worthy of more detailed investiga-
tion in different regions of the world.
CONCLUSION
The landscape of microfinance has changed dramatically, especially at the

upper end, where capital markets are rapidly becoming integrated with the
financing and investment needs of the microfinance markets. This rapid
growth and inflows of comparatively large amounts of capital presents its
own set of challenges. On the other hand, at the lower end, NGOs remain
the mainstay for the poorest of borrowers, where the loan sizes are very
small and the interest rates remain high. Technological innovations may
22 Microfinance
well hold the key to reducing the costs of delivering small loans and accept-
ing very small savings deposits. They may in turn pave the way to making
the microfinance approach more scalable. The regulatory challenges asso-
ciated with the rapid change in the landscape deserve greater attention from
policy-makers and researchers.
An issue that the book does not consider but is perhaps extremely
important is the fact that the poor lack very basic services that the rest of
society takes for granted. Such services include: (1) access to primary edu-
cation, (2) access to primary health care, and (3) access to and training in
operating more modern tools and technology in day-to-day activities.
Access to these is as important if not more so than access to financial ser-
vices. On the other hand, it is clear that the access to financial services will
enhance the ability of poor households to access these important basic ser-
vices and skills as were. A number of practitioners have already recognized
the need to deliver in parallel both these basic services and the financial
services.
NOTES
1. World Bank, PovertyNet website: http://go.worldbank.org/33CTPSVDC0.

2. Akerlof, G.A. (1970), “The market for ‘lemons’: quality uncertainty and the market
mechanism”, The Quarterly Journal of Economics, 84 (3), 488–500; Stiglitz, J. and A.
Weiss (1981), “Credit rationing in markets with imperfect information”, American
Economic Review, 71 (3), 351–66.
3. Flannery, M. and K. Samolyk (2005), “Payday lending: do the costs justify the price?”,
FDIC Center for Financial Research working paper, Chicago University, Chicago.
4. Caskey, John P. (1991), “Pawnbroking in America: the economics of a forgotten credit
market”, Journal of Money, Credit and Banking, 23 (1), 85–99.
5. See Global Envision (14 April 2006), “The history of microfinance”, www.globalen
vision.org/library/4/1051/.
6. Many nationalized banks in India extended credit to “priority sectors” such as agriculture.
In Indonesia, Bank Rakyat Indonesia (BRI) extended micro-savings and credit products.
7. The motivation for this initiative underscores the view of many financial institutions:
“We subscribe to the view that microfinance investing can contribute to a double bottom
line,” said Ed Grzybowski, TIAA-CREF’s Chief Investment Officer, “GMIP, and this
investment in ProCredit, gives us an opportunity to seek competitive returns through
socially responsible investments that we believe have a low correlation to traditional
equity and fixed income markets,” quoted in “TIAA-CREF launches $100m
microfinance program”, in SustainableBusiness.com (2006), accessed 18 October at
www.sustainablebusiness.com./index/cfm/go/news.feature/id/1384.
8. Gonzalez, A. (2007), “Efficiency drivers of microfinance institutions: the case of oper-
ating costs”, MicroBanking Bulletin, 15 (Autumn).
9. CGAP (2007), “CGAP reflections on the Compartamos IPO: a case study on
microfinance interest rates and profits”, Focus Note, 42 (June).
10. Gonzalez, A. (2007), “Resilience of microfinance to national macroeconomic events: a
look at MFI asset quality”, MicroBanking Bulletin, 14 (Spring).
11. The criteria used for the target group classification and the participating institutions are
detailed in Gonzalez (2007) (see Note 10). Median values are reported in the tables.
12. Local moneylenders, who operate with limited capital, provide the outside borrowing
options for the poor, besides friends, relatives, and trade credit from local shops.
Anecdotal and scattered evidence suggests that the interest rates charged by the money-
lenders are significantly higher than the rates under the microfinance alternative.
13. See also Zaman, S. and S.N. Kairy (2007), “Building domestic capital markets: BRAC’s
AAA securitization”, MicroBanking Bulletin, 14 (Spring).
14. Reddy, R. (2007), “Microfinance cracking the capital markets II”, Insight, 22 (May).
15. See Note 9.
16. CGAP (2008) “Microfinance capital markets update”, 23 (January), www.cgap.org/
mcm/archives/V23_0108.html.
17. Stephens, B. (2007), “Commercialization continues apace”, MicroBanking Bulletin, 14
(Spring).
18. Hishigsuren, G. (2006), “Transformation of microfinance operations from NGO to
regulated MFI”, IDEAS, www.microfinancegateway.org/content/article/detail/36733.
19. Pearlstine, J. Boon (2006), “Lenders, borrowers hook up over the web: Prosper.com and
other sites provide forum for individual bidders willing to offer small loans”, Wall Street
Journal, 20 May; also see Credit Union Magazine, January 2008.
20. Zopa now has operations in the United States.
21. Current websites of these organizations may be found at www.kiva.org/; www.prosper.com;
www.lendingclub.com/home.action; https://us.zopa.com/.
22. Milbrandt, J. (2008), “Biometrics and smart cards serve as successful microfinance inno-
vations in Asia”, Microfinance Report, 9 January.
23. www.cgap.org/policy/branchlessbanking.
24. Kumar, R. (2004), “eChoupals: a study on the financial sustainability of village Internet
centers in rural Madhya Pradesh”, Information Technology and International
Development, 2 (Fall), 45–73.
25. Milbrandt, J. (2008), “The rise of mobile phone banking”, Microfinance Report, 17
January; Chemonics (2006), “Mobile-phone banking expands into rural Philippines”, 24
May, www.chemonics.com.
26. Global Envision (2007), “Microfinance goes mobile: cell phone banking revolutionizes
financial services for the poor”, 3 August, www.globalenvision.org/library/4/1708.
27. Shylendra, H.S. (2006), “Microfinance institutions in Andhra Pradesh: crisis and diag-
nosis”, Economic and Political Weekly, 20 May; Microcapital (2006), “Microfinance
institutions reach crucial agreement with government in Andhra Pradesh, India”,
11 October, www.apmas.org/pdf%5Cn.pdf; Microcapital (2006), “Indian Bank – ICICI
reaches arrangement with provincial government on micro-loan interest rates”,
www.microcapital.org/?p=580.
28. Shylendra, H.S. (2006) as Note 27.
29. This episode also called attention to the potential for increased default risk, and the polit-
ical-economic nature of microfinance. The Andhra Pradesh state government had taken
a tough stance following allegations that usurious interest rates and heavy-handed loan
recovery procedures contributed to farmer suicides. Four lending institutions—
Spandana, Asmita, Umduma Poddu Pedatha, and SHARE Microfin came in for exten-
sive scrutiny. In addition, ICICI Bank, which writes most of these loans using these MFIs
as disbursement and collection agents, faced potential write-offs worth $100 million.
30. Some microfinance institutions require savings as a condition for lending, and impose
costs for providing such services. Products such as savings, insurance, and loans are
bundled in a way that the overall costs to the consumer can become prohibitive. See
Shylendra (2006), note 27.
31. “Branchless banking: rapid growth poses regulatory challenges,” CGAP Focus Note,
accessed 31 January, 2008 at www.cgap.org/p/site/c/template.rc/1.26.2154.
32. The analysis is based on MIX database, which is in the public domain. There is no entry
in Table 1.5 for credit unions and cooperatives for East Asia as there was only one such
institution there during the sample period considered. MIX data only included institu-
tions that voluntarily report their financial and outreach data, and hence is necessarily a
biased and under-represented sample of the population of the lending institutions and
24 Microfinance
borrowers in reality. We nevertheless believe that the MIX data can provide very useful
stylized facts about the issues studied in this book.
33. The sample size of reporting institutions is as follows: the data have 37 reporting banks,
124 cooperatives and credit unions, 158 non-bank financial institutions, and 295 NGOs.
The data are not necessarily contemporaneous and may have a time lag of up to one year.
34. Stiglitz, J.E. (1990), “Peer monitoring in credit markets”, World Bank Economic Review,
4 (3), 351–66.
35. In Pakistan, of the six non-bank financial institutions reporting, the percentage of
women borrowers ranged from 4.1 percent to 34 percent. On the other hand, two NGOs
from Pakistan report a near 100 percent empowerment in favor of women! In
Afghanistan, of the three non-bank financial institutions reporting, the percentage
of women borrowers ranged from 45 percent to 65 percent. One NGO reports 100
percent women borrowers and the other two report 25 percent and 40 percent participa-
tion by women.
2. The role of international capital
markets in microfinance
Brad Swanson
INTRODUCTION
In 2004, international capital markets awoke to the attractiveness of invest-

ing in microfinance. Since then, debt and equity security issues for
microfinance have raised an estimated US$1 billion from private sector
financial institutions seeking commercial returns.1 The deals have taken
forms that are familiar in developed markets such as initial public offerings,
bond issues, collateralized debt obligations (CDOs), and securitizations of
the underlying micro-loans. In addition, private sector debt and equity
microfinance funds have sprung up—for investors who prefer to give dis-
cretion to professional managers—and are now thought to control more
than $2 billion, of which more than $300 million is “mainstream” com-
mercial investment.2
Overall, cross-border investment in microfinance surged to $1.4 billion in
2006, triple the rate only two years previously.3 While traditional suppliers
of microfinance capital—non-profit organizations, governmental develop-
ment agencies, and individuals—are contributing to this surge,4 the novelty
since 2004 is the participation by private sector institutional investors
seeking full market returns. These mainstream commercial investors, most
located in Western Europe and the USA,5 are driving the opening of capital
markets to microfinance.
How and why commercial mainstream investors have come into
microfinance and the likely evolution of capital markets funding for
microfinance is the topic of this chapter.6
THE NEED FOR CAPITAL MARKETS FUNDING IN

MICROFINANCE
When properly conducted, microfinance is a profitable, low-risk, and

expanding financial activity. For example, from January to June 2007, the
25
26 Microfinance
26 widely dispersed microfinance institutions (MFIs) in Microfinance

Securities XXEB, a $60 million CDO sponsored by Developing World
Markets (DWM), had an aggregate annualized return on equity of more
than 25 percent and were growing their loan portfolios by more than 50
percent on an annual basis, while their “PAR-30” (total amount of “port-
folio at risk,” or loans with payment delays, beyond 30 days) was only 2.9
percent.7 This is a performance that any commercial bank would be proud
to announce.
Already, the number of borrowers served by MFIs is globally estimated
at 100 million.8 With an average loan size of $170, the total market size is
estimated at $17 billion. Yet the potential demand is 15 times the current
market—estimated at 1.5 billion, or half the 3 billion global working poor.
Thus, microfinance represents a total commercial market of more than
$250 billion.
Currently more than three-quarters of the $17 billion funding total is
raised from domestic markets. However, this number is skewed by the
amount—almost $8 billion—coming from deposits in the few countries
where MFIs are allowed to take deposits. Most of the estimated 10 000
existing MFIs are not deposit-taking institutions—and are unlikely to
become so, given the cost and complexity of complying with regulations
typically applied to institutions taking deposits from the public. Future
funding for MFIs is thus unlikely to be sourced mainly from deposits.
Domestic emerging country commercial banks, which should be major
funding sources for MFIs, are typically averse to lending to them (see the
section on “Local Currency” below). Moreover, capital markets in most
developing countries are thin and the major institutional players are averse
to or legally constrained from significant investment in microfinance. For
these reasons, it is unlikely that domestic sources in emerging countries will
generate more than a fraction of the more than $200 billion that will need
to be raised to satisfy potential demand.
Moreover, while non-commercial investors account for 80 percent of the
$4 billion in funding now sourced internationally, this is a legacy of the
origin of microfinance in charitable and officially sponsored development
activity. As MFIs’ appetite for capital grows exponentially, it is unlikely
that government agencies and non-profit organizations will increase their
flow of funding proportionately: first, they will be faced with competing
demands for assistance; and, second, they will begin to question whether
their mission is best served by funding financial enterprises that are
profitable and are increasingly transforming into privately owned com-
panies able to attract commercial investment. (However, this realization
may not have begun to sink in yet—see discussion of role reversal below in
the section on “The Contribution of Non-commercial Investors.”)
The role of international capital markets 27
The only available source of funding for commercial lending of this mag-
nitude is the international capital markets. Already, microfinance invest-
ment vehicles, which typically include private sector institutional investors,
are growing their investment portfolios at 233 percent per year, while official
development agencies are lagging at 150 percent.9 For the international
capital markets, funding a $200 billion industry is routine.
FROM FUND TO CDO
The first10 microfinance fund to reach beyond socially responsible investors

was established in 1998. The Dexia Microcredit Fund, sponsored by Dexia,
a Franco-Belgian bank, and advised by BlueOrchard Finance SA, based in
Geneva, offers investors a return above their cost of funds (typically 1–2
percent over a benchmark rate11) and an ability to redeem their investments.
In November 2007, funds under management were $233 million.12
As a fund (a Luxembourg-based SICAV13) offering redemption rights to
investors, Dexia needs to keep its maturities to MFIs relatively short and a
large portion of its assets in cash (typically 20 percent or more). This limits
returns to investors and the attractiveness of Dexia’s funding to MFIs,
many of which need longer-term maturities on a portion of their liabilities
to better manage risk.
In 2004, after six years of operations and with $45 million under man-
agement, BlueOrchard wanted to provide longer-term funding to MFIs
and more attractive rates to investors. It partnered with DWM, an emerg-
ing markets fund manager and advisor based in Connecticut, to create the
first CDO in the microfinance industry. In this transaction, loans were
made to MFIs for seven years from the proceeds of issuing fixed rate bonds.
As the bond investors were not entitled to their principal until the bonds’
maturity, there was no need to keep large quantities of cash on hand to deal
with redemptions. Furthermore, MFIs had use of the funds for the full
period with no interest rate uncertainty.
The CDO was named BlueOrchard Microfinance Securities I (BOMSI).
The first closing of $40 million occurred in July 2004 and a subsequent
closing of $47 million was held in April 2005.
This transaction looked very different from any existing microfinance
investment vehicle and it marked the beginning of mainstream capital
markets investment in microfinance. The major innovations in micro-
finance funding pioneered by BOMSI include the following:
1. BOMSI is not a fund—investment decisions are not handed off to

a professional manager. There is no asset substitution or active
28 Microfinance
management. Investors in BOMSI have a single source of repayment,

a static pool of 14 loans to MFIs taken on at closing. When investors
came into BOMSI, they did so on the basis of their own assessment of
the credit risk of the underlying MFIs—and they have to live with this
decision for seven years. Legally, BOMSI is a special-purpose (legal)
vehicle (SPV)—a limited liability corporation—registered in the busi-
ness-friendly state of Delaware. The vehicle is limited by its constitu-
tional documents solely to servicing its loans to MFIs and repaying its
creditors. Cashflows from debtors to creditors pass transparently
through the vehicle. When the loans pay off and the liabilities mature,
BOMSI will make its final payments to investors and be liquidated (see
Figure 2.1).
2. BOMSI’s funding is stratified in five levels of risk—senior, three classes
of subordinated, and, at the bottom, equity. (Both BlueOrchard and
DWM are equity investors in BOMSI.) The cashflow from BOMSI’s
loans to MFIs is applied according to a strict order of precedence,
known in structured finance as the “cash waterfall.” Senior investors
are paid completely first, then the other classes in order of precedence.
Equity investors do not get a current return on their investment but if,
after all MFI loans have reached maturity and all other investors have
been repaid, there is residual cash left in the BOMSI SPV it will be allo-
cated to the equity investors.
3. BOMSI’s investors do not hold units in a fund and have not made loans
to BOMSI. Rather, they have purchased securities—bonds and equity
interests. As we will see later, this distinction was important in attract-
ing institutional investment.
Investors
Debt service
1st priority
BOMSI Senior
Special-
purpose 2nd priority Subordinated
legal vehicle (three classes
A,B,C)
MFIs 3rd priority Equity
Trustee Servicer Advisor
Service providers
Figure 2.1 Cashflows from loan repayments
These elements are common to CDOs and other forms of securitization in

more developed asset classes such as mortgages, corporate loans, auto
loans, or student loans. But these are asset classes with substantial data
going back a number of years describing default performance under a
number of economic scenarios. In the microfinance industry, by contrast,
MFI write-off policies vary widely and data on micro-loan defaults typi-
cally are not recorded consistently by different MFIs. Moreover, these data
typically are neither independently audited nor rigorously modeled to
determine likely performance under varying circumstances. (Although
recently, a non-profit research firm, Center for the Development of Social
Finance, did a static pool analysis of more than 600 000 micro-loans from
two MFIs—SKS in India and IMON in Tajikistan—using developed world
methodology, in order to demonstrate that at least some MFIs are rigorous
enough in their record-keeping to permit this style of analysis.)14
Moreover, BOMSI securitized loans to only 14 institutions in nine
countries—much less diversification than typical CDOs or other securit-
ization transactions in developed markets, where the asset pool may com-
prise many hundreds or thousands of loans.
Given these factors, implementing a CDO for the microfinance industry
required changing the way investors viewed both microfinance and the
CDO product.
INTRODUCING COMMERCIAL INVESTORS TO THE

MICROFINANCE CDO
Despite the relative paucity of data and diversification, DWM, which took
primary responsibility for structuring the transaction, encouraged investors
to compare BOMSI to mainstream commercial investments. DWM held the
view that to attract sufficient investor interest, BOMSI had to reach beyond
the circle of funders primarily motivated by social, not financial, returns.
To distinguish BOMSI as a commercial investment—different from
investment funds, donations to NGOs, or other means then available to
support microfinance—DWM highlighted the following:
● Low default rate in MFI loan portfolios. All participating MFIs

reported default rates below 1 percent. Although reporting systems
were not consistent or their results independently verified, the pro-
fessionalism and the track record of the MFIs themselves added
credibility to their findings.
● Favorable risk/return ratios. The tiered capital structure enabled
BOMSI to offer high returns to the higher-risk tranche investors,
30 Microfinance
while providing the lower-risk investors with a substantial degree of

collateralization, enabling them to feel satisfied with a low credit
spread over the benchmark Treasury bond because their notes had
the highest priority of repayment. Investors were not asked to dis-
count their return expectations in view of the presumed social value
of microfinance. With a variety of securities offering different risk
and return parameters, DWM was able to segment the international
investor base and thus appeal to a wide spectrum of potential
investors.
● Familiar investment instruments. BOMSI debt investors purchased
bonds drafted in their language, and carrying features common to
commercial bonds. They benefited from the appointment of a trustee
to safeguard their interests, as is the case in most bond issues. The
bonds are transferable and each series is endowed with a unique
CUSIP15 number that facilitates record-keeping, valuation, and per-
mitted transfers. (However, the bonds were privately placed, are not
listed, and are not intended to be actively traded.) These features
helped to ensure that investors had a high comfort level with the form
of the investment and could focus clearly on the underlying risk and
return.
In one important respect BOMSI was differently structured from other

commercial transactions: Overseas Private Investors Corporation (OPIC),
a US government development agency, purchased the most senior tranche
of securities. Note that OPIC’s ownership of the senior tranche conveyed
no protection to more junior investors—by virtue of the cash waterfall,
they were exposed to risk in the MFI loan portfolio ahead of OPIC.
However, the participation by a large and well-respected development
agency—often referred to as “the halo effect”—encouraged investors who
otherwise might have been unwilling to consider the transaction.
GROWING PARTICIPATION BY COMMERCIAL

INVESTORS
In the event, the first closing of BOMSI attracted only $1.5 million, or 4
percent of the capital raised, from private sector investors seeking a full
market return (see Table 2.1 below). However, by the time of the second
closing, nine months later in April 2006, interest in the transaction had
spread and commercially motivated institutional investors accounted for 41
percent of the amount invested. Moreover, the commercial investment
came from a wider spread of investor types.
Table 2.1 Commercially motivated CDO investors by type and risk

category
Investor type BOMS 1 BOMS 2 MFS Total

USD USD USD USD
Bank 500 000 9 139 640 9 639 640
Money manager 1 000 000 500 000 3 036 000 4 536 000
Insurance company 500 000 1 000 000 1 500 000
Pension fund 18 000 000 20 500 000 38 500 000
University endowment 100 000 100 000
Total 1 500 000 19 100 000 33 675 640 54 275 640
% of total investment 4% 41% 56% 37%
Risk category
Equity 1 500 000 1 500 000
Juniora 500 000 600 000 125 320 1 225 320
Mezzanineb 1 000 000 18 500 000 1 125 320 20 625 320
Senior 30 925 000 30 925 000
Total 1 500 000 19 100 000 33 675 640 54 275 640
Notes:
a. For BOMS 1 and 2, subordinated notes C and B.
b. For BOMS 1 and 2, subordinated notes A.
Source: DWM.
A little over a year later, in June 2006, DWM closed its third CDO trans-
action, Microfinance Securities XXEB (MFS), for which it was sole
sponsor. This $60 million securitization of loans to 26 MFIs had more
investment primarily commercially motivated than primarily socially moti-
vated. Moreover, for the first time commercial investors (besides the
sponsor) purchased equity. By this time, not only had market familiarity
with microfinance grown, but DWM had also obtained an investment
grade rating—A—on the MFS senior notes from MicroRate, a specialized
microfinance rating agency. This heightened commercial investors’ comfort
with the senior tranche. In addition, DWM had sponsored a study indicat-
ing that microfinance is less correlated to economic downturn than other
emerging markets assets, making portfolios including microfinance, in
theory, less volatile (see the section below, “On the Path to an Asset Class”).
This development was of interest to commercially motivated investors.
Table 2.1 shows the amount of investment in three CDO transactions
contributed by institutional investors seeking full market returns, with
socially positive impact a desirable additional benefit. The remainder of the
32 Microfinance
investment came from investors whose primary motivation was social—

thus, for these, financial return was of secondary importance.
High net worth individuals (HNWIs) constituted 10 percent of the
investment amount in the first BOMSI close. (They are not shown in Table
2.1 as we do not characterize them as commercial investors.) This percent-
age fell in the second close and by the closing of MFS, HNWIs as a group
were down to under 5 percent of the total capital invested.
While there is doubtless a significant potential market among HNWIs,
and among retail investors generally, for microfinance risk, financial insti-
tutions are the bellwethers as they have greater sophistication, more
resources, and stronger tolerance for volatility and illiquidity.
The market for microfinance CDOs has continued to grow. Notably,
BlueOrchard has sponsored two more CDOs, with Morgan Stanley as
placement agent, in April 2006 and May 2007, raising $99 million and $110
million respectively. The entrance of Morgan Stanley, a “bulge bracket”
investment bank, is another signal that microfinance funding is gaining
credibility as a capital markets activity.
However, in the latter half of 2007, and continuing into 2008, the failure
of a number of CDOs based on sub-prime mortgages in the USA made the
very term “CDO” suspect in the eyes of many institutional investors and
slowed the pace of growth in microfinance CDOs, even though the two
types of assets are unrelated. As MFIs in microfinance CDOs have contin-
ued to perform well, capital markets intermediaries believe that receptivity
among investors to this asset class will improve with increasing recognition
of the inherent robustness of microfinance credit risk.
The relatively small volumes outstanding to date and the legal restric-
tions on trading privately placed securities mean that secondary markets
have not developed. Secondary markets should not be anticipated until the
number of participants and amounts outstanding increase significantly,
including several listed issues to act as price indicators for the markets.
CDOs VS. FUNDS
CDOs were the first non-fund capital markets products in microfinance for
several reasons:
● MFIs typically have balance sheets that are too small to justify trans-
actions of the scale required to access international capital markets—
aggregating MFI loans is necessary.
● On the other hand, MFIs are used to borrowing internationally—cre-
ating loans to international standards and packaging them into the
asset side of a special-purpose financing vehicle does not present

insuperable challenges.
● Capital markets investors predominantly demand instruments
denominated in USD or euros (although the decline of the dollar and
relative stability of many emerging market currencies in recent years
are persuading investors to become more open to local currency
risk). On the other hand, MFIs typically (but not always) lend in local
currency; therefore, they are used to borrowing in hard currency and
passing the risk on to their clients, whose demand for loans is rela-
tively interest rate inelastic.
● Top-quality MFIs are found throughout the emerging markets so
geographic diversification can be achieved.
● MFIs typically have very few loan products, so their risk on the asset
side is relatively easy to analyze and can serve as a proxy for the
underlying risk of the micro-borrower—highly diversified, highly
granular—that strongly draws the typical capital markets investor.
Despite these factors, the relative scarcity of top-quality MFIs may act
to brake the growth of this asset category. Of an estimated 10 000 MFIs
worldwide, fewer than 100 have qualified for inclusion in a CDO to date. As
market demand for CDOs grows, CDO arrangers will need to push farther
“down the pyramid” to tap MFIs of lesser size and credit quality to gener-
ate assets. But, given the absence of data in the microfinance industry, as
noted above, the analysis of risk in CDOs is not a function of statistics but
rather of individual assessment of MFIs.
Investors find it difficult to make the time necessary to take individual credit
decisions on numerous MFIs, especially given that the investment represents
only a very small part of the investor’s portfolio responsibility. Up to now, the
presence in CDOs of MFIs that are mostly top-ranked—demonstrated either
through ratings or performance over time—has served to ease these credit
decisions. But with the top tier of MFIs growing “overbanked” (see the
section below, “Is Microfinance Riding for a Fall?”), CDO arrangers will need
to persuade investors to take risks on MFIs that are less known or appear
financially weaker. Part of this persuasion may come through education—
some smaller MFIs may be as creditworthy as their larger peers—but struc-
tural features such as credit guarantees or higher collateralization levels may
become necessary in some deals to assuage investor concern.
While the CDO has broken new ground as an investment instrument in
microfinance, investment funds have also been growing, and, as previously
noted, are thought today to control more than $2 billion of capital. Of
course, investment funds in microfinance are not new. Traditionally, low-
return or no-return funds sponsored by non-profit organizations have been
34 Microfinance
a major source of funding for microfinance. What is new is an emphasis on

funds that actually offer a return to investors.
Even as of January 2008, of the 89 microfinance funds listed by
MicroCapital, a microfinance news and research service, only 26 are char-
acterized as actually seeking a financial return.16
Would-be institutional microfinance fund managers have several hurdles
to overcome in persuading clients to invest:
● Investment fee “cascades”. Institutions that manage funds make it a

practice for their funds not to invest in other funds in order to avoid
a build-up of fees that erodes the ultimate returns to their investors.
Also, ceding investment discretion to others may appear to weaken
their own standing as managers.
● Lack of transparency. Investors may find it difficult to understand the
pricing, volatility, and performance of assets that exist primarily in
funds, as the portfolio effects and the manager’s screening activities
could mask the underlying data.
● Liquidity. Funds typically trade off liquidity (that is, redemption) for
return. If they provide an easy exit for investors, funds that invest in
illiquid assets like microfinance will find it necessary to keep a rela-
tively large percentage of their portfolio in low-yielding cash. Investors
with long time horizons, as many institutions have, may prefer to invest
directly in the underlying assets and run the liquidity risk.
However, funds do play an important role in the growth of capital markets

access for microfinance. For example, many institutions will choose a fund
as their first investment in a new asset category, relying on the manager’s
experience and knowledge of the market to enhance the investor’s comfort
level, as well as to gain familiarity with a multiplicity of MFIs through a
single investment.
MICRO-LOAN SECURITIZATIONS
CDOs and funds that specialize in senior loans to MFIs are not the only
capital markets instruments in microfinance. Direct securitization of micro-
loans has attracted a great deal of interest, as micro-loans are relatively
homogeneous and vastly diversified. As the spectacular growth in recent
years of asset-backed securities in international markets makes clear,
investors welcome a “pure play” risk on granular financial assets. However,
several important constraints are slowing the emergence of a true asset-
backed notes product in microfinance:
● Short maturity of micro-loans. As opposed to 30-year mortgages,

most micro-loans mature in less than a year and feature frequent
amortization, so that all but the shortest-term micro-loan securitiza-
tions will need to incorporate a mechanism to roll over or substitute
the underlying assets, which greatly increases the structuring com-
plexity and administrative cost.
● Origination risk. Because the portfolio of underlying micro-loans
needs constant replenishment, the ability of the MFIs to originate
continually a sufficient volume of micro-loans is a significant ad-
ditional risk.
● Important role of servicer. Successful MFIs cultivate intimate rela-
tionships with borrowers. Thus, the MFI role in servicing securitized
micro-loans is a critical element in the performance of the securitized
portfolio. This makes it difficult to portray micro-loan securitizations
as pure borrower risk. In effect, the performance risk of the MFI ser-
vicer is a key component in the overall risk profile—and a difficult
one to quantify, much less hedge against.
● Government regulation. Many emerging market jurisdictions have
non-existent, rudimentary, or inflexible regulatory structures that
pose daunting obstacles to the legal structuring necessary to set up
securitization vehicles, execute true sales of micro-loans, and trans-
fer payments transparently to offshore investors.
Given these constraints, there have been only two cases of micro-loan secu-
ritization in international capital markets (as opposed to CDOs that secu-
ritize loans to MFIs), and both of them have featured substantial credit
enhancement by non-commercial investors. First in May 2006, ProCredit
Bank Bulgaria, a subsidiary of ProCredit Holding AG, sold €48 million of
its loan portfolio to institutional investors in a deal rated BBB by Fitch
Ratings. The European Investment Bank and KfW, the German develop-
ment agency, provided partial guarantees.17 Four months later, BRAC, a
large Bangladesh MFI, held the first close of a program, backed by micro-
loans, which will issue $15 million (local currency equivalent) of six-month
maturity notes twice a year for six years. The issue was rated AAA by a local
rating agency. The partial guarantors were KfW and the Dutch develop-
ment agency FMO.18
EQUITY
As MFIs mature and transform from non-profit organizations into com-

panies, including in some cases regulated institutions, their need for equity
36 Microfinance
grows. With the high return on equity and fast growth of the industry, the
internal rate of return of MFI equity investment looks compelling on
paper. Consequently, at least 15 private equity funds mobilizing $620
million (much of it from non-commercial sources, however), have been set
up to address this need.19
The major uncertainty in commercial equity investment in MFIs is the
small number of “exits,” that is, portfolio investment liquidations, to date.
Most private equity investors look more to capital gains upon sale of their
stakes and less to dividends as the principal component of their return. This
is appropriate in microfinance as MFIs need to retain earnings in the busi-
ness to finance further growth if they are to escape an endless cycle of
sourcing fresh equity. But without a deep track record of successful exits,
the private equity investor is entitled to puzzlement if not skepticism
regarding the prospective return on MFI equity investment.
The only private equity fund that has gone through a complete cycle of
investment and liquidation is ProFund Internacional SA, which from
1995–2005 invested approximately $20 million total in ten Latin American
MFIs for an annual average return of 6 percent. ProFund is of interest here
not for its financial returns—it was sponsored by socially motivated
investors and did not set out to maximize profits—but for its success in re-
alizing all ten exits within its allotted ten-year life.20
All but one of ProFund’s exits came from sales to shareholders or spon-
sors of portfolio MFIs, several of them pursuant to a put (that is, a con-
tract requiring one counterparty to purchase an asset at a specified price
from another party at the seller’s option) or pursuant to an agreement
among existing shareholders. While effective in the case of ProFund, exits
to insiders (management, major shareholders, and sponsors) are worrisome
to private equity investors if they are the only feasible means of liquidating
investments. Investors prefer a mix of mechanisms including those that
bring in third-party buyers, such as initial public offerings (IPOs), mergers,
and acquisitions, in order to set arm’s-length pricing and foster competi-
tion. Moreover, puts to insiders expose the put-holders (that is, investors)
to the credit risk of put-writers (that is, insiders), and expose the put-writers
to substantial future liabilities they may not be willing to take on, or may
accept only at very conservative valuations. If a put can be agreed, and the
credit risk of the counterparty is acceptable, the risk-adjusted return is not
likely to excite the private equity investor.
Acquisitions by financial or strategic investors are more welcome path-
ways to exit, but there have been very few examples of this in microfinance.
Microfinance networks might seem to be likely acquirers but most, whether
for-profit or non-profit, prefer to build their own operations in new coun-
tries from the ground up or to partner with smaller, non-corporate MFIs.
No substantial organization has attempted a “roll-up,” or a growth strat-

egy through acquisition to date.
IPOs have been used to provide exits to investors in two significant
cases—Equity Bank in Kenya and Compartamos in Mexico. The latter
transaction, in April 2007, garnered much publicity, some of it unfavorable,
for putting $450 million into the hands of existing investors who had paid
approximately $2 million for these shares originally.21 Equity Bank also
rewarded its early investors, but on a smaller scale. These examples have
given MFI owners and private equity investors hope that IPOs will provide
lucrative exit opportunities. However, few emerging country stock markets
have sufficient liquidity to provide assurance of full valuation. In addition,
both Compartamos and Equity Bank are relative giants in their jurisdic-
tions. As market leaders and first-movers, they represented unique invest-
ment opportunities that by definition later entrants to these domestic
public markets will not provide.
It is more likely that MFI acquisitions will provide consistent exit paths
for private equity investors. Strategic investors such as commercial banks,
leasing companies, and insurance companies will see the value in MFIs not
just as lenders but as delivery vehicles for other financial services to a pro-
prietary and loyal customer base. These potential investors in many cases
will come cross-border, recognizing that the fundamentals of micro-lending
are roughly similar in most countries, as shown by the success of networks
that apply a common methodology across the developing world. Already
some Western European banks have purchased Eastern European banks
that specialize in small and micro-enterprise lending in order to extend their
footprint into the European Union hinterland.
Another likely source of acquisition is by a competitor. Already some
countries, including Bolivia, Ecuador, India, Nicaragua, and Peru, are
seeing competition among MFIs that previously relied for growth on an
under-penetrated market.
THE CONTRIBUTION OF NON-COMMERCIAL

INVESTORS
As we saw in the BOMSI case, an official development agency can provide

credibility and ease market acceptance of a product even without direct
enhancement of risk. But as international capital markets grow more famil-
iar with microfinance, the value of the “halo effect” is diminishing. Yet non-
commercial investors are not superfluous in microfinance. They can play a
valuable role in taking on risks that commercial investors don’t understand
or are uncomfortable with, and in so doing, leverage this investment. For
38 Microfinance
example, the Global Commercial Microfinance Facility (GCMF), spon-

sored and managed by Deutsche Bank, is a $75 million fund whose
investors include socially responsible HNWIs, official development agen-
cies (from the USA, the UK, and France), foundations, and also a number
of commercially motivated investors such as banks, insurance companies,
and pension funds. The facility is designed to make it easier for MFIs to
obtain local currency loans from local banks.22 While we deal with local
currency issues later in this chapter, the importance of the facility for this
section is to recognize that most institutional investors are uncomfortable
taking local currency risk, especially inasmuch as many currencies in
emerging markets either cannot be hedged or can only be hedged at unac-
ceptable cost.
In essence, the GCMF takes advantage of the ability of non-commercial
investors to shield commercial investors from risks they are unwilling to
take on, thus leveraging the risk capital of the non-commercial investors to
the benefit of both.
Another, less obvious, example of the catalytic role of non-commercial
investors is the $11.4 million bond issued by Microfinance Bank of
Azerbaijan (MFBA) in August 2007, managed by DWM. This was the first
case of a bond issued in international capital markets by an MFI without
credit enhancement. While MFBA had a growing and profitable business,
the issue’s attractiveness was bolstered by investors’ perception that the
AAA-rated development agencies owning a majority of MFBA shares
—European Bank for Reconstruction and Development, International
Finance Corporation, and European Investment Bank—would step in if
the issuer faced financial difficulty rather than face the embarrassment of
a default in a portfolio investment.
Whether providing a halo to comfort commercial investors or actually
taking on risk that commercial investors feel uncomfortable with, non-
commercial investors can significantly speed up access to capital markets
investment for MFIs. But it appears that bilateral and multilateral devel-
opment agencies are going beyond this role and are actually crowding out
private sector investors in commercially credible deals.
MicroRate, a Washington, DC-based MFI rating agency, has published
a study23 claiming that “development agencies are today heavily concen-
trating their funding on the largest and most successful MFIs, exactly the
target investment market of private investors.” The study posits that devel-
opment agencies tend to make easy choices and are squeezing private
investors out of the market with their subsidized finance rates.
In 2005 (last full year of data), the study found that the development
agencies increased their direct funding to top-rated MFIs by 88 percent. At
the bottom of the pyramid, where MFIs are most in need of the “patient
capital” and technical assistance that these agencies provide at taxpayer

expense, the development agencies actually cut their funding to the lowest-
rated MFIs by 25 percent.
Shortly after the appearance of the publication, a number of private
sector microfinance funders joined together to appeal to the development
agencies to change this practice, but the results have been inconclusive.
Clearly, if development agencies see their roles as competing with private
sector investors, they will slow the access of microfinance to capital
markets.
LOCAL CURRENCY
One of the largest constraints to growth of microfinance funding is the illi-

quidity and volatility of many local currencies in the developing world. Of
course, if MFIs were able to rely on local funding sources, this would not
be a problem. But, as we noted earlier, the bond markets of most develop-
ing countries are thin and poorly regulated. Moreover, institutional
investors, the largest capital sources in these countries, are often highly
restricted in their permitted range of investments.
Paradoxically, local commercial banks, which should be a major source
of funding for MFIs, in many countries are less likely to accept MFI risk
than foreign banks. This is symptomatic of the larger problem of risk-
aversion among these banks. In many countries, capital-hungry governments
crowd out private lender borrowers. In some countries, banks are content to
lend to large corporations, state-owned entities, and foreign businesses and
are under no pressure to expand their presence into smaller indigenous busi-
nesses. In some countries, banks have simply not made the effort to under-
stand and analyze MFI risk, assuming that “banking the unbankable,”
whether directly or indirectly through MFIs, cannot be prudent.
Foreign investors typically are uncomfortable with local currency risk
that cannot be hedged. This means that many MFIs must borrow in dollars
or euros and push the risk onto their borrowers. Fortunately for the MFIs
the short maturities of their loans gives them flexibility to effectively re-
price their assets to account for currency fluctuations. Even more fortu-
nately for the MFIs, most borrowers are unable to access capital from other
sources and so accept interest rate hikes that a more affluent and competi-
tive market would challenge. Nevertheless, adjusting constantly to unfore-
seeable shifts in exchange rates is a strain on MFI operations and imposes
additional risk on borrowers.
On occasion, MFIs and offshore lenders hedge by depositing the hard
currency loan in a local commercial bank, which then lends to the MFI in
40 Microfinance
local currency, secured by the deposit. (In a variant of this technique, the
deposit-taking bank is different from the local bank but issues the local
bank a standby letter of credit to secure the risk of the MFI local currency
loan.) Although the local bank’s loan to the MFI is effectively risk-free, the
local bank frequently will not reduce the interest rate to the MFI by a large
enough quantum so that the combination of the local currency interest rate
plus the guarantee fee paid to the offshore lender for taking the risk works
out as a feasible financing cost for the MFI.
A number of initiatives are underway to provide unorthodox hedging
facilities for capital markets investors in thinly traded currencies. The
Dutch development agency FMO, for example, is putting together a swap
vehicle capitalized with $350 million in equity that would support $1.5
billion outstanding in currency swaps that are beyond the maturity avail-
able commercially. By acting as swap counterparty for a basket of emerg-
ing market currencies, the facility aims to achieve risk mitigation through
diversification while providing a substantial return to equity investors.24
Ultimately, local currency markets will mature and provide efficient and
flexible hedging tools. In addition, by that time, local capital markets may
have sufficiently matured to lessen the strain put on foreign investment to
meet MFIs’ growing capital needs.
ON THE PATH TO AN ASSET CLASS
The term “asset class” has a number of definitions. From an institutional

investor’s standpoint, an asset class is a kind of asset that is suitable for
inclusion in an investment portfolio. In order to be suitable, the asset class
must fulfill certain requirements. Fundamentally, it must be recognizable as
a distinct kind of asset, such that different investments in the same asset
class can be analyzed together, can substitute for each other, and can be
relied upon to perform similarly in similar circumstances.
Crucially, the asset must be liquid, so that portfolio managers can trade
into and out of the asset easily according to their changing viewpoint and
their portfolio’s cashflow. Liquidity is a function of several factors includ-
ing volume, exchange listings, ratings, research, and so on.
Additionally, it is important that the asset has a track record, data that
can be analyzed to make predictions about price changes in response to
market conditions. If the asset is relatively less correlated to other assets in
the portfolio, that is, of course, a positive as the overall volatility of the
portfolio will be reduced by including the new asset in the mix.
Overall, microfinance funding is a long way from meeting these require-
ments. It approaches the definition most closely in its distinctiveness and
relative homogeneity. But it is extremely illiquid and likely to remain so for

an extended period of time while volumes build up. Secondary markets are
not likely to develop until there is a critical mass of exposure among a large
number of investors so that willing buyers can be matched with willing
sellers.
Interestingly, a case can be made that microfinance is largely uncorre-
lated to other emerging market assets and so would reduce portfolio volatil-
ity, or beta. In a study sponsored by DWM and carried out by New York
University25 the operating performance under different economic scenarios
of 283 MFIs in 65 developing countries was compared with that of 112
commercial banks from 33 developing countries. The findings were that
MFI financial results are less sensitive to economic downturn than that of
emerging market commercial banks. While the authors concede that the
study is based on somewhat inconsistent and incomplete data, it neverthe-
less serves as a useful indicator and will likely lead to further useful inves-
tigation of the characteristics of microfinance as a prospective asset class.
IS MICROFINANCE RIDING FOR A FALL?
Looking ahead, some microfinance investors see events on the horizon that
worry them:
How will microfinance perform in turbulent economic conditions? The

concern is that the risk of default in the event of global or even localized
recession is unknowable, and may be substantial, for MFIs that have only
operated during periods of prosperity.
This fear overlooks the fact that microfinance as a financial service
segment is not nascent, even though capital markets only recently “discov-
ered” the asset. Many MFIs have been in business for 10–20 years and have
weathered significant economic and political instability in countries such as
Indonesia and Bolivia. Experience and research, such as the correlation
study noted earlier, indicate that MFIs are inherently less vulnerable to eco-
nomic shocks than other finance providers. (Of course, a sovereign event
such as rescheduling or capital controls, or a breakdown of law and order,
could force default on even the strongest and most liquid MFI, as well as
any other debtor to external markets.)
The top tier of MFIs shortly may be “overbanked” The fear is that too
much investment is chasing too little opportunity and that returns are
falling to the extent that investors will lend imprudently to lower-quality
MFIs in order to meet return expectations. The current compression of
42 Microfinance
emerging market spreads relative to higher-rated paper, while cyclical, high-

lights this concern.
However, while many of the best-known and largest MFIs are attractive
candidates for investment, many smaller and more obscure MFIs also have
high-quality credit risk. This stems from the underlying robustness of the
microfinance business model.
Most micro-enterprises operate “under the radar” of the formal
economy. The level of economic activity they engage in is so basic as to be
immune from the normal ebb and flow of the economic and political
systems they operate in. Their operating margins are commonly quite high
(although, of course, small in absolute terms). Their employees are family
members or close associates whose terms of employment are informal and
flexible. Their owners’ liability for business debts is not limited by a legal
form—micro-borrowers take personal responsibility for the loans made to
them, and they know their ability to continue to make a living, and often
to maintain the respect of their community, is intrinsically tied to their
punctual payment of all amounts due.
For the MFI, administering the loan book is time-consuming and labor-
intensive, but once the procedures are carefully designed, inculcated, and
tested in practice, operations are usually stable, and extending the customer
base of the MFI by opening new branches becomes almost routine.
Financial controls need to be strict and minutely observed, however.
In fact, it is difficult to find instances of default by MFIs that seek self-
sufficiency (that is, do not view themselves as charitable operations) and
have been in business several years. Certainly some MFIs may have sought
support from the international networks they belong to in order to shore
up a weakened balance sheet or improve faulty operations. In addition,
some MFIs are believed to understate their portfolio at risk numbers by
routinely extending the maturity of overdue loans. But MFIs that practice
this usually do end up collecting close to 100 percent of the principal and
interest from the overdue borrowers.
Are MFIs abandoning their core constituency? A third concern is the move
of some MFIs “up market” along with their more successful clients. While
the vast bulk of MFI activity currently consists of small loans to individ-
ual micro-entrepreneurs some MFIs have begun to offer more sophisticated
services to larger clients involving more substantial risks—small business
lending, mortgages, factoring, leasing, insurance, and so on—and also
enhanced revenue. Generally speaking, a larger loan is more profitable to a
financial institution than a smaller one, as administration costs do not
increase proportionately with loan size. This is a controversial develop-
ment. Some observers denounce MFI “mission drift” and worry that MFIs
will abandon their low-income clients as they progress upstream. Others

believe MFIs can continue to remain committed to poverty alleviation and
still retain their more successful clients as they accumulate wealth.
As these products take on more importance on MFIs’ balance sheets, the
analysis of the MFIs’ financial strength will grow more complicated, and their
performance vis-à-vis other emerging markets assets may grow more highly
correlated, reducing their value in lowering portfolio beta. On the other hand,
as these MFIs grow to more resemble mainstream financial institutions, both
in terms of size and structure, they may attract the attention of some main-
stream analysts, traders, and investors, further enhancing investment sources
and liquidity. Ultimately, while some MFIs may turn their backs on their
origins, most will keep their focus on micro-loans even while providing
higher-level services, both because microfinance is good business in itself and
because it will provide the breeding ground for the higher-value customers.
CONCLUSION
The rush of capital markets investment in microfinance is unprecedented

and it is wise to question its sustainability. Certainly, risks to continued
growth abound, and we have noted a number of them, including:
● eventual exhaustion of investment opportunities at the well-known

and accessible tip of the MFI pyramid;
● structural obstacles to providing investors with direct exposure to
micro-loans via securitization;
● scarce track record of equity exits;
● lack of clarity regarding the role of non-commercial investors;
● underdeveloped local capital markets, coupled with insufficient
hedging tools for foreign currency investment;
● illiquidity, sparse data, and small volumes slowing the journey
toward achievement of “asset class” status;
● “mission drift” eroding MFIs’ distinctive risks and returns, and less-
ening their value in reducing portfolio volatility.
Many of these risks reflect the fact that microfinance has only recently been
introduced to capital markets. They should ease over time as investors accu-
mulate exposure to this asset. Extrapolating current trends points to
financial products that are more numerous, more standardized, and more
fitted to capital markets norms. At the same time, secondary markets will
make their appearance, and ratings agencies and researchers (both com-
mercial and academic) will focus more attention on the sector. Specialized
44 Microfinance
hedging tools will ease the distortions of too much lending in foreign cur-
rency. These developments should abet liquidity and help to give investors
comfort that microfinance is suitable for regular allocations of portfolio
investment. In effect, investor demand for assets itself will become an
important and self-fulfilling driver of progress in microfinance.
Moreover, as MFI owners and managers grow accustomed to an envi-
ronment in which a deep pool of commercial funding is available for the
well-run, expanding MFI, we can expect strategic transactions—mergers,
acquisitions, buy-outs, roll-outs, listings, and so on—to become integral
elements in the life cycle of successful MFIs. This will result overall in
stronger, more efficient, and more skilled institutions better serving clients’
needs.
Of course, too rapid growth could also lead to speculation, overheating,
and a crash, as we have seen many times before in financial markets, from
junk bonds to high tech to sub-prime. And certainly some MFIs will
expand too quickly and lose control of their costs and their loan books, or
cut rates too aggressively for competitive reasons, or push their clients into
over-indebtedness. Microfinance is no more immune to excess than any
other business activity. But the inherent robustness of the microfinance
business model lays down a strong foundation for solid growth, and the
sizable potential market ensures absorption capacity for substantial fresh
financing.
Overall, the distinctive focus of microfinance on “banking the unbank-
able”—bringing financial services to customers outside the formal financial
system—gives it a unique and attractive profile of risk and reward that can
draw institutional investors seeking diversification and absolute return—
even those who are unmoved by the prospect of promoting social values.
NOTES
1. Estimate by the author.

2. Littlefield, E. (2007), “Building financial systems for the poor”, presented at Cracking
the Capital Markets Conference, New York, 19 March. The author adjusted the data to
remove commercial investment in CDOs (which are treated in this chapter as transac-
tions, not funds) and to add more recent funds aimed at the commercial market, such as
the €160 million SNS Institutional Microfinance Fund, managed by DWM, which had
its first closing in May 2007.
3. CGAP (2007), “Microfinance investment vehicles”, CGAP Brief, April, www.cgap.org/
portal/site/CGAP/menuitem.95cb370f4995240167808010591010a0/.
4. For example, the CGAP brief cited in Note 3 states that portfolio investment by inter-
national financial institutions doubled from 2004 to 2006, while microfinance investment
channels targeted to individuals report increasing assets, such as Kiva (www.kiva.org).
See Walker, R. (2008), “Extra helping”, The New York Times Magazine, 27 January.
5. DWM research.
6. In this chapter, “capital markets” means transactions or funds in which all or a major
portion of the investment is raised from private sector institutional investors seeking
fully risk-adjusted returns.
7. DWM research.
8. Data in this paragraph and the next three come from “Optimizing capital supply in
support of microfinance industry growth”, a presentation by McKinsey & Company to
the Microfinance Investor Roundtable in Washington, DC on 24 October 2006. The
exception is the reference to 10 000 MFIs, an estimate widely quoted in the literature; see
for example, Odell, Anne Moore (2008), “Microfinance: catch the swelling SRI wave”,
Sustainability Investment News, 11 January.
9. Littlefield (2007), p. 3—see note 2.
10. According to the Dexia website, 27 January 2008, www.dexia.com/e/discover/
sustainable_funds 2.php.
11. BlueOrchard uses a standard short-term rate as its benchmark, six-month LIBOR
(London Inter-bank Offered Rate), the rate that prime banks charge each other for
liquidity.
12. Dexia Micro-Credit Fund Monthly Newsletter, BlueOrchard Finance, November
2007, www.blueorchard.org/jahia/Jahia/Site/blueorchard/Products/pid/42/dexiannews
letter.
13. A SICAV (société d’investissement à capital variable) is an open-ended fund common in
Western Europe especially Luxembourg, Switzerland, Italy, and France, comparable to
a mutual fund in the USA.
14. CDSF (2007), “Capital markets-style risk assessment: testing static pool analysis
on microfinance”, Center for the Development of Social Finance, March,
www.cdsofi.org/downloads/MFIStudy-CDSF-Mar 07.pdf.
15. CUSIP is an acronym for the Committee on Uniform Securities and Identification
Procedures, a standards body. A CUSIP number uniquely identifies a specific security to
facilitate custody and trading of securities.
16. MicroCapital, 27 January 2008, www.microcapital.org/?page_id=7.
17. “Press Release”, ProCredit Holding AG and Deutsche Bank, 15 May 2006.
18. Rahman, R. and S. Shah Mohammed (2007), “BRAC micro credit securitization series
I: lessons from the world’s first Micro-credit backed security (MCBS)”, MF Analytics,
Ltd, Boston, 20 March, www.microfinancegateway.com/files/45785_file_11.pdf.
19. DiLeo, P. and D. FitzHerbert (2007), “The investment opportunity in microfinance”,
Grassroots Capital Management LLC, June, www.grayghostfind.com/industry_
insights/viewpoints/the_investment_opportunity_in_microfinance.
20. “ProFund Internacional, SA (2008)” www.calmeadow.com/profund.htm.
21. Rosenberg, R. (2007), “CGAP reflections on the Compartamos initial public offering: a
case study in microfinance interest rates and profits”, Focus Note (42), June.
22. USAID (2007), “The Deutsche Bank global commercial microfinance consortium and
USAID’s DCA guarantee”, United States Agency for International Development,
January, www.microlinks.org/ev_en.php?ID=17450_201&ID2=DO_TOPIC.
23. Abrams, J. and D. von Stauffenberg (2007), “Role reversal: are public development insti-
tutions crowding out private investment in microfinance?”, MicroRate, February,
www.microrate.com/pdf/rolereversal.pdf.
24. “TCX–the currency exchange” (2008), www.fmo.nl/smartsite.dws?id=88.
25. Krauss, N. and I. Walter (2006), “Can microfinance reduce portfolio volatility?”, Stern
School of Business, New York University, November.
3. Securitization and micro-credit
backed securities (MCBS)
Ray Rahman and Saif Shah Mohammed
INTRODUCTION
BRAC Micro Credit Securitization Series I, closed in August 2006, was

the world’s first securitization of micro-credit receivables and the first of
a new type of investment called micro-credit backed security (MCBS).
This securitization was also the first AAA-rated transaction within
Bangladesh.
A number of innovative transactions have taken place in the micro-credit
industry in the last few years. In 2004, ICICI bank purchased 25 percent of
SHARE Microfin Ltd’s micro-loans in a US$4.9 million transaction. The
purchased microfinance receivables were valued at their net present value at
an agreed-upon interest rate. A portion of the transaction was guaranteed
by Grameen Foundation USA. Also in 2004, BlueOrchard issued its first
collateralized debt obligation (CDO). In these, and subsequent transac-
tions, BlueOrchard pooled collections of loans made out to a number of
microfinance institutions around the world.
Unlike the ICICI transaction, the BRAC securitization is scalable and
allows for tranching. It is not a one-time sale of receivables at a discount.
Unlike the BlueOrchard CDOs, the BRAC securitization involves the direct
pooling of micro-credit receivables instead of the pooling of loans disbursed
to microfinance institutions. The investments are thus directly linked to the
performance of the underlying portfolio of micro-credit receivables instead
of the risk of the originating institution. The transaction is entirely in local
currency—thus removing any currency risk from the originating institu-
tion. As such, the BRAC securitization represents a step in the evolution of
the linking of microfinance to capital markets.
In general terms, the transaction size is US$180 million in local
Bangladesh currency. It consists of 12 equal tranches with the asset pool
backing each tranche, mirroring the overall risk profile of BRAC’s micro-
credit portfolio. The tenor of the transaction is 6.5 years, with each tranche
maturing in 12 months. It is a fixed rate private placement sold pre-
46
Securitization and micro-credit backed securities 47
dominantly to local Bangladesh investors. BRAC is the servicer, and

Eastern Bank Limited (EBL) is the trustee.
Having arranged and structured this transaction, we discuss the story
behind it, and lessons from it for the future. The next section deals
with the transaction rationale that convinced BRAC to agree to going
forward with a securitization. The third section examines various politi-
cal economy considerations that had to go into arranging the transaction.
The fourth looks at the actual structure of the transaction. It highlights
various data and logistical constraints that were reflected in the final struc-
ture. Working with BRAC’s micro-credit portfolio data allowed us to spot
certain risks that were not anticipated at the beginning of the structuring
process, and these risks were mitigated in the final structure. Entering the
eighth month of the transaction, we are able to draw some lessons
about how well the structure has held, and what can be taken from it in
future transactions. The fifth section thus examines the performance of
the bond and the underlying pool of securitized micro-credit receivables
since the beginning of the transaction. We scrutinize the various credit
enhancements that were put in place, and explore which of them could be
reduced or removed in future securitizations. The final section, as the con-
clusion, relates some final lessons from the transaction and ideas for the
future.
TRANSACTION RATIONALE
Established in 1972, BRAC is Bangladesh’s largest NGO, with nearly

100 000 employees. BRAC takes a holistic approach to development, com-
bining micro-credit activities with health care, education, and social advo-
cacy. With its 1381 area offices, 35 000 primary schools and 35 health
centers, BRAC has a presence in every district in Bangladesh.
As of March 2007, BRAC’s micro-credit program had more than 5.3
million borrowers with outstanding loans. BRAC’s micro-credit activities
are carried out through three programs targeted at different borrower
groups:
1. Dabi—with loans ranging from around US$50 to US$500 in size, tar-

geting individuals with less than 1 acre of land;
2. Unnati—with loans ranging from around US$150 to US$850 in size,
targeting individuals with more than 1 acre of land, or individuals who
have graduated from the Dabi group;
3. Progati—with loans ranging from US$350 to US$5000 in size, target-
ing small enterprises.
48 Microfinance
Table 3.1 BRAC microfinance sources of funding 2002–05
Source (%) 2002 2003 2004 2005 CAGRa (three-year)

Members’ savings 36.70 45.70 48.33 46.30 8.05
Loan from PKSF
(quasi-government) 28.50 23.85 15.71 8.14 –34.14
Grant from donors 16.30 12.97 10.35 8.53 –19.42
Retained earnings 11.90 14.05 15.02 16.34 11.15
Loan from banks
(including
syndication) 6.10 3.24 9.50 20.69 50.25
Others 0.00 0.00 1.09 0.00
Note: a. Compound annual growth rate.
Source: BRAC.
Loans generally have a one-year maturity. The interest rate is a flat 15

percent, with collections taken on a weekly basis for Dabi and Unnati, and
on a monthly basis for Progati.
The BRAC portfolio has been growing rapidly in the last few years.
Between January 2006 and January 2007, for example, the principal out-
standing on loans grew by 35 percent from US$280 million to more than
US$380 million, and the number of borrowers grew by more than 10
percent to more than 5 million borrowers.
BRAC has been funding its micro-credit activities from a number of
sources, including the savings of its group members/borrowers, donor
grants, and funding from the government and some donor agencies
through loans from Palli Karma Sahayak Foundation or PKSF, the apex
microfinance institution. In early 2004, BRAC borrowed the equivalent of
nearly US$30 million from the local capital markets through its first syndi-
cation. Table 3.1 gives the breakdown of BRAC’s microfinance funding
sources prior to the transaction.
These financing options had severe drawbacks. Donor funding was
volatile, and could not be relied upon in long-term planning. BRAC also
felt itself under increasing pressure from the government to reduce the
interest rates charged to borrowers as a precondition for further loans
through PKSF. However, BRAC considered its current interest rates nec-
essary for the sustainability and viability of its micro-credit program and
felt that they could not be reduced further. Financing through syndications
was problematic in the long run as interest rates on future syndications
would be driven by the volatile credit market and would not adequately
reflect the historical performance of BRAC’s micro-credit program.

Besides, syndications could not be carried out indefinitely without falling
foul of regulatory limits on borrowing.
Securitization presented itself as an attractive alternative, or at least com-
plement, to the existing sources of funding. Securitization would allow
BRAC to have a more efficient balance sheet. It would improve BRAC’s
asset/liability management, and reduce leverage. Properly structured, the
transaction could obtain a higher rating than BRAC, and allow BRAC to
raise lower-cost funds. A longer-term transaction would give BRAC the
ability to plan out its explosive growth. And significantly, BRAC would
have a broader investor base. Through the transaction, BRAC would be
able to remove its dependence upon PKSF funding, thereby alleviating
some of the pressure to reduce interest rates.
POLITICAL ECONOMY CONSIDERATIONS
The Approval Process
Bangladesh Bank
The central bank initially welcomed the concept of securitization as a way
to deepen the capital markets. In December 2004, around the time the
BRAC transaction was initially discussed, Bangladesh Bank arranged a
well-attended workshop on securitizations as a source for funding for infra-
structure and development projects. And yet, over the next two years
Bangladesh Bank did little to actually implement many of the proposals
that came out of the December 2004 workshop. For whatever reason, the
bank simply refused follow-up on many of the changes one might deem nec-
essary for having more efficient capital markets. For example, in the work-
shop (and in other public and private discussions with the bank), the need
for a yield-curve, was noted by the local financial community. As we write
this report, the yield-curve is still non-existent in the Bangladeshi banking
sector, and there are no official efforts in place to create such a benchmark.
Our own discussions with Bangladesh Bank about securitizing micro-
credit receivables were well received in the months after December 2004.
And based on these interactions with Bangladesh Bank, the structure for
the transaction was created and submitted to the bank for approval around
June/July 2005. The original signals from the bank indicated that the
approval process would be a short one. After all, the structure addressed the
concerns that it had raised, namely currency risk concerns and the involve-
ment of local financial institutions. The entire currency risk of the foreign
investment was borne by FMO of the Netherlands, and two-thirds of the
50 Microfinance
transaction would be subscribed by local investors. Yet, Bangladesh Bank

withheld approval until the end of December 2005/January 2006.
Bangladesh Bank raised three objections. First, we would have to
further reduce the involvement of foreign investors by reducing the size of
the tranche covered by a KfW/FMO guarantee. Second, the guarantees
would have to be removed completely after the first year. Third,
Bangladesh Bank was worried about income from the bonds and fees
being remitted outside of Bangladesh by FMO and KfW. This concern was
directly related to the overheating of the local US dollar market in the
second half of 2005, which saw the exchange rate move up in a matter of
months from around taka 60/USD to more than taka 70/USD. Due to the
increase in global oil prices, the government-owned Bangladesh Petroleum
Corporation borrowed US dollars heavily from nationalized commercial
banks.
The structure was changed to meet the first two concerns. For the third
concern, we were able to show that the impact on the local dollar market
from remittances would not be as strong as anticipated by Bangladesh
Bank. The structure required FMO to buy a new tranche every six months,
and this would actually help in the local currency crisis. Further, the regu-
lators were simply an order of magnitude off in their understanding of the
size of fees and other remittances.
Securities and Exchange Commission

The approval process at the Securities and Exchange Commission (SEC)
was similarly fraught with delays. Since the bond issued in each tranche
would have a one-year maturity under our structure, the existing rules did
not require an SEC approval. The SEC confirmed this in our conver-
sations with it while the transaction was being structured. However, the
Bangladesh Bank approval letter required us to approach the SEC again
on this question. In mid-January 2006, we approached the SEC for a final
confirmation that the approval process was not required under the rules.
The SEC now decided that it would require us to get approval after all
because of the six-year term of the transaction, even though the securi-
ties issued themselves would be of one-year maturity. Then in early
February 2006, we were again surprised with a long request for docu-
ments and analyses. We submitted these documents and analyses almost
immediately. In March 2006, we received an AAA rating for the transac-
tion from the Moody’s-affiliated Credit Rating Agency of Bangladesh
(CRAB). After this, the SEC did not logically have a reason to withhold
approval. However, it was only in June 2006 that we received the
approval.
What Explains the Delays in the Approval Process?
We can only speculate about the reasons behind the delays in the approval
process. Some of the concerns expressed by the regulators—such as pres-
sures on the local dollar market—were genuinely felt. And regulators may
very well have been uncomfortable with the transaction due to their unfam-
iliarity with both microfinance and with securitizations. After all, this was
a first-of-a-kind transaction with a number of moving parts (as described
in the next section). Further, the bureaucratic regime is not one known for
its transparency or nimbleness. Based on informal discussions, we can
suggest that some of the delays may have been related to deeper concerns
of the government.
PKSF
Established in 1990, PKSF is the apex micro-credit organization in
Bangladesh, a private–public partnership between the microfinance NGOs
and the government. The organization receives funding from the govern-
ment, the International Development Association (IDA)/World Bank,
USAID, the Asian Development Bank (ADB), and the International Fund
for Agricultural Development (IFAD). It lends these funds to partner
microfinance organizations at a below-market rate.
PKSF has successfully utilized its funding clout to induce microfinance
organizations to strengthen their reporting and auditing processes, as well
as implement computerized management information systems (MIS) at the
head office (and in some cases, area office) level, thereby strengthening insti-
tutional capacity. The organization has played some role in coordinating
the activities of microfinance organizations to better target groups that may
not have been receiving microfinance services. It has also facilitated
research on microfinance activities.
BRAC itself had received US$30–40 million in loans from PKSF. But as
discussed previously, BRAC felt itself under increased pressure to reduce the
interest rates it charged as a precondition for being able to borrow again
once the PKSF loans were paid down. One of the rationales for the securi-
tization was thus to move away from dependence on PKSF funding.
Further, PKSF may have felt its role as the apex microfinance body in
Bangladesh threatened by the prospect of a microfinance organization being
able to raise funds through securitization without PKSF involvement.
The government
2005 was the UN International Year of Microcredit. In what is arguably the
home of micro-credit, the government’s reception of the year was schizo-
phrenic. On the one hand, the government sponsored a number of
52 Microfinance
workshops and events to celebrate the year and highlight the achievements
of microfinance institutions. On the other, the headlines in Dhaka were at
times dominated by statements from the Minister of Finance publicly
doubting whether the activities of microfinance institutions had any posi-
tive role to play in development. In the last quarter of 2005, there was a
public disagreement between the Chairperson of BRAC and the Minister
of Finance over the issue of interest rate caps.
But the government’s apathy to microfinance activity in Bangladesh was
not animated solely by policy concerns about high interest rates. Since
2001, the centre-right BNP had ruled Bangladesh in alliance with the right-
wing Jamaat-i-Islami. Jamaat represented many of the most conservative
elements of Bangladeshi society, and historically the relationship between
these constituents of Jamaat and the activities of the development NGOs
has been a tense one. Further, the NGOs have been vocal about the rights
of minorities under the BNP–Jamaat coalition. The influence of Jamaat on
government policy since 2001 has been widely reported.
A less well-publicized aspect of the relationship between the government
and the microfinance sector in Bangladesh has been concerns about the
involvement of microfinance institutions in political activities. With mil-
lions of borrowers and beneficiaries of health care and education pro-
grams, microfinance institutions have an unprecedented ability to mobilize
voters. In past years, prominent organizations such as Proshika have been
accused of openly partisan activities.
The Proshika case is worth looking at closely to get a sense of govern-
ment concerns. Like BRAC, Proshika was involved in education and social
advocacy along with its microfinance activities. Like BRAC, Proshika had
also been able to scale up its microfinance activities to reach millions of
borrowers in nearly every part of Bangladesh. In the mid-1990s, Proshika
played a prominent role in the opposition Awami League-led movement for
elections under a neutral caretaker government. In the 1996 and 2001 elec-
tions, BNP accused Proshika of slanting its voter education material in
favor of the Awami League. And in April 2004, Proshika was accused of
trying to mobilize thousands of its borrowers to assemble in the capital to
launch an opposition platform. In May 2004, the BNP/Jamaat government
cracked down upon Proshika, arresting its chairperson and stopping the
flow of donor funding to its programs.
Whatever the merits of the Proshika case, government misgivings about
the influence of the large microfinance institutions and their potential for
quickly mobilizing thousands for political causes led in 2004 to the draft-
ing of laws deepening government control of NGOs. It was in this climate
that the BRAC securitization was proposed to the regulators. Concerns
about the perception of Proshika’s activities had led BRAC to form the
Federation of NGOs of Bangladesh (FNB) with other NGOs, breaking

away from the Proshika-led Association of Development Agencies in
Bangladesh (ADAB), the apex development group. But the government
likely felt that BRAC’s dependence upon the whims of government funding
through PKSF was a vital lever for controlling its activities.
STRUCTURE OVERVIEW
The Structure
The parties
BRAC is the originator in this transaction. BRAC also plays the role of ser-
vicer, depositing the collections from the securitized receivables and collat-
eral to the special-purpose vehicle’s (SPV’s) accounts on a monthly basis,
updating the pool of securitized receivables and collateral, and reporting
the performance of the pool to the investors. The trustee for the SPV is
Eastern Bank Limited—a leading local bank.
The investors in this transaction are FMO, Citibank, and two leading
local banks, Pubali Bank and The City Bank. Citibank’s investment shall
(for the first year) be guaranteed by FMO and counter-guaranteed by
KfW of Germany. As in most securitizations, BRAC is also the residual
beneficiary of all cash flows after fees, principal payments, and interest pay-
ments are paid out each month by the SPV.
MF Analytics shall provide continuing support to BRAC to maintain
the securitized pool of receivables and collateral and assist in reporting
performance to the investors. RSA Capital was the lead arranger of the
transaction. FMO, KfW, and Citibank were co-lead arrangers.
The transaction was rated by the Credit Rating Agency of Bangladesh,
the local Moody’s affiliate. BRAC’s MIS system and MF Analytics’ pooling
and reporting algorithms were audited by PriceWaterhouseCoopers.
BRAC itself is audited by Ernst & Young Malaysia.
Tranche structure
This US$180 million transaction is divided into 12 equal tranches over six
years. Every six months, the originator shall sell US$15 million worth of
micro-credit receivables to the trust/SPV created for this transaction in return
for cash. It shall also earmark another US$7.5 million worth of micro-credit
receivables as additional collateral. The trust shall issue US$15 million worth
of certificates or bonds of one-year maturity backed by the pool of securi-
tized receivables and collateral to the investors. In return, the investors shall
invest US$15 million in the trust on or before the date of issue for the tranche.
54 Microfinance
Each tranche is divided into three sub-tranches. Sub-tranche A is the

FMO investment, amounting to US$5 million. Sub-tranche B is the
Citibank investment, also US$5 million. As noted already, Sub-tranche B
is guaranteed by FMO for the first two tranches, and counter-guaranteed
by KfW. Sub-tranche C, the remaining US$5 million, is issued to the local
investors. The certificates issued to the investors in each sub-tranche are
pari passu (equitable). It should be noted that the cost of funds for BRAC
from this transaction is between 150 to 200 basis points below what would
have been available to it had it gone for a straight loan or syndication.
The securitized asset pool

The underlying asset pool for the securitization consists entirely of BRAC’s
micro-credit receivables. At the outset of each tranche, the pool or the col-
lateral underlying the tranche is selected to reflect the distribution of loans
in BRAC’s current portfolio (excluding all loans that have liens from PKSF
and syndications) along three variables:
1. program (Dabi/Progati/Unnati);
2. geography (identified by area offices);
3. activity or purpose of the loan (identified by a “schematic code”).
The securitized loans are selected within these categories randomly. As a result,
the pool is not biased towards including loans of a particular size or age.
The pool is over-collateralized by 50 percent, that is, at the beginning of
each tranche, US$22.5 million (in equivalent Bangladesh taka) of receiv-
ables are pooled as per the criteria noted above. Furthermore, the asset pool
is replenished with additional collateral from month to month if the fore-
casted cash flow from the pool is less than 140 percent of the SPV liability
in the following month.
Bond pay-down structure for each tranche

The certificates or bonds issued for each tranche are of one-year maturity.
The bonds are amortized monthly based on a predefined pay-down sched-
ule. Interest is also paid on the outstanding principal outstanding of the
bonds on a monthly basis. The pay-down schedules of the certificates reflects
the actual pay-down of the underlying securitized receivables. The principal
pay-down schedules for the first two tranches is presented in Table 3.2.
With new tranches issued every six months over a course of six years,
BRAC is provided a committed, long-term source of funding. The tranches
disburse funds at a rate that BRAC can absorb without trouble for distri-
bution to its borrowers. The one-year maturity of the bonds reflects the
short-term nature of the underlying micro-credit receivables, allows the
Table 3.2 Principal pay-down schedules for tranche 1 and tranche 2
Montha % Principal Paid Down

Tranche 1 Tranche 2
1 12.25 12.000
2 12.25 12.000
3 12.25 12.000
4 12.25 12.000
5 12.00 11.750
6 9.85 11.500
7 9.00 9.750
8 7.50 7.500
9 5.90 4.500
10 4.00 4.250
11 1.90 2.000
12 0.85 0.750
Note: a. September 2007–February 2008.
securitized pool to track BRAC’s portfolio, and provides the investors an

additional level of comfort for investing in this first-of-a-kind transaction.
Credit enhancements
A number of credit enhancements were included in this transaction. The 50
percent over-collateralization of the securitized pool and the replenishment of
the pool with additional receivables in the event of projected cash flows falling
below 140 percent of the following month’s SPV liability, have already been dis-
cussed. A few additional credit enhancements were added for good measure:
Substitution Loans identified as delinquent and loans with missing or cor-

rupted data are removed on a monthly basis from the pool. (The delinquent
loans are under our definition of delinquency. Loans identified as defaulted
loans under BRAC’s definition are not removed from the securitized pool.)
In their place, loans from the same program, geographical region (area
office), and type of activity as far as possible are purchased from BRAC as
replacements by the SPV. These replacement loans are selected so that they
would mature later than delinquent and missing data loans.
Removing prepayment risks Instead of being used to pay the investors

or released to the residual beneficiary, prepayments are captured in a
provisional account. Once loans with prepayments within the securitized
pool mature, these prepayments are released at the rate at which these
56 Microfinance
amounts would have been collected had no prepayment taken place.

Prepayment risk is thus totally hedged out from the transaction.
Debt service reserve account A debt service reserve account (DSRA) of

US$2.5 million was provided by BRAC as an additional credit enhance-
ment at the beginning of the first two tranches. The DSRA amount is
expected to be negotiated down after tranche 2.
MF Analytics pool maintenance and reporting modules

The structure required the creation of a software package by MF Analytics
tailored to this transaction. The package included modules that create the
pool of securitized receivables, and forecast the cash flow from the securitized
pools and bond pay-down structure at the beginning of each tranche. It also
included modules that on a monthly basis substitute delinquent and missing
loans, trap and release prepayments to completely hedge out prepayment
risk, and replenish the pool if necessary with additional collateral. Additional
modules fulfill all of the reporting needs of BRAC for this transaction.
Issues
The need for disaggregating data

The investors desired the securitized pool at the time of pooling to reflect
the risk characteristics of BRAC’s entire micro-credit portfolio. However,
once the pooling took place, the portfolio would not track the evolution of
BRAC’s portfolio. Rather, because of substitution, prepayment trapping,
and replenishment, the pool would have its own risk characteristics. The
risk characteristics of the pool would thus diverge from those of BRAC’s
portfolios over the course of the transaction. Further, as already noted, the
substitution of loans on a monthly basis and the tracking and controlled
release of prepayments above would require the processing of individual
loans. Thus, it simply would not do to analyze BRAC’s historical portfolio
data at an aggregated level for structuring.
The structuring of the transaction required two levels of data analysis:
1. We had to identify the drivers of risk in the available data, and design the
selection process of the securitized pool so the distribution of identified
characteristics in BRAC’s portfolio would be reflected in the selected
pool. This involved identifying and analyzing the qualitative and quan-
titative data available on the loans and borrowers in BRAC’s portfolio.
2. Once the pool was selected, we had to run simulations of the dynamic
pool on the available historical data as well as on simulated data to
reflect various stress conditions to test and refine the structure.
Disaggregating the loan data required us to process and analyze enormous

amounts of data. At the time of the structuring, BRAC had nearly 5 million
borrowers, and the transactions of these loans were tracked monthly. Each
securitized pool contained between 250 000 and 300 000 loans.
Our analyses of BRAC’s information system and portfolio also revealed
a number of issues that needed to be mitigated in the final structure:
Information and logistical issues
Time lag BRAC updates the collection and disbursement data at its 1381
area office computers daily. However, these data are transferred to the Dhaka
head office only once a month. The infrastructure simply does not currently
exist for more frequent transfers of information to the head office. At the
head office, it takes around a week to complete the process of updating and
checking the database. (At the time of the structuring, this process took
nearly 10–12 days.) As a result, the longest gap between a transaction in the
field and the information reaching the head office is nearly 42 days. A delin-
quency or prepayment will in many cases be reported to the head office nearly
42 days after they took place. The structure would have to “solve” this lag.
Changing collection dates BRAC’s loans to its borrowers are collected on

a weekly or monthly schedule. However, the exact collection dates for a par-
ticular loan cannot be known until after the end of the month for a number
of reasons.
No collections take place on local holidays, and these holidays are often
dependent upon the sighting of the moon. BRAC’s system will, during the
month, allow the collection schedules to be updated to reflect these local
holidays. Similarly, at a country-wide level, national holidays—many of
them based on the lunar calendar—require shifts in collection dates.
Furthermore, changes in BRAC’s personnel itself can lead to changes in
collection dates. At BRAC, each collection officer is assigned two village
groups a day to meet borrowers, examine their activities, collect payments,
and make new disbursements. The transfer or promotion of a collection
officer can result in a change in the day of a week that a particular village
group will be visited.
Forecasts of collections from the pool of securitized receivables would
have to reflect such uncertainty about collection schedules.
Missing, inconsistent, and unusable data The monthly data transferred to the
head office arrives in the form of CDs or zip drives carried by couriers from
the 1381 area offices. These data are uploaded to the head office servers before
any reporting can take place. To say that the transfer process is not fail-safe is
58 Microfinance
an understatement. Each month, BRAC’s head office finds missing informa-

tion and corrupted data, and has to ask individual area offices to transfer the
information again. But such checks do not spot all missing information.
At each of the area offices, the information for thousands of collections
and disbursements are entered each day by a BRAC-trained computer
operator hired solely for this purpose. While the software used by BRAC
does incorporate some checks on whether information was entered cor-
rectly, some mistakes inevitably creep in. As a result, BRAC’s data included
some inconsistent information.
Further, the software used for data entry at BRAC itself generated some
inconsistencies. We were able to spot some of these inconsistencies in our
due diligence process, and BRAC corrected them.
Finally, at the field offices, there is no way to fix malfunctioning com-
puters without sending the machines to Dhaka. Sometimes malfunctioning
machines are not spotted early, and some of the information transferred to
the head office is corrupt and unusable. Because of fires, theft, or natural
disasters, there is always some underlying risk of spoilage of data.
Notwithstanding these constraints, we found that errors and inconsist-
encies in BRAC’s dataset were relatively rare, less than 1 percent of the
data every month. We also noticed that over time, BRAC’s monthly dataset
got better, for a number of reasons. BRAC removed the software-generated
inconsistencies that we spotted. Its internal checks on data quality
improved. And the process of transferring data became more streamlined.
However, the structure would have to take into account the risk of some
missing, corrupt, or inconsistent data.
Data issues in the risk analysis process
Unavailability of historical data One of the limitations that we had to deal

with in the structuring process was the incompleteness of the historical
information available to us. BRAC area offices had been computerized at
different times, and it was only near the end of 2005 that the computeriz-
ation process was completed. Around the same time, BRAC was also in the
midst of upgrading the data-entry and database software used at the area
offices. While accessible computerized data for some area offices went back
a few years, for most area offices accessible data went back only a few
months. And because the upgrades to the area office systems had not been
completed, at the time of structuring the deal we were only able to analyze
individual loans on a monthly basis, instead of being able to look at the
actual dates of transactions. Thus, we would only be able to know if a par-
ticular loan failed to pay in a particular month, but not the exact date when
it failed to pay.
Absence of data on borrower characteristics The BRAC data available to

us at the time of the structuring process included information on where the
loan was being used and what the loan was being used for. But important
demographic characteristics were missing, such as the estimated age of the
borrower, the length of time they had been borrowing from BRAC, infor-
mation on defaults on previous loans, and even the number of installments
missed in the current loan. These constraints severely limited our ability to
create credit scores for individual borrowers.
While the upgrades to BRAC’s system allow some of this information to
be captured, BRAC’s database still does not include useful information that
could be easily incorporated. BRAC’s health care and education programs,
for example, are a rich source of information on borrower households. Yet
these databases still do not speak to each other.
Deciphering the meaning of defaults At the beginning of the transaction,

there was some inconsistency in the way that BRAC and the investors
understood what constituted a defaulting loan. BRAC defines a defaulting
loan as one that failed to pay its total obligation of principal and interest
by the end of the one-year period of the loan. Thus, a loan that missed all
weekly payments for months but was still to reach the end of its one-year
term is considered a “current loan” by BRAC. For BRAC, it is this
definition that generates the astonishing, well-publicized nearly 100 percent
repayment rates.
Such a definition was simply not palatable for the investors. But a
problem in creating a definition of default or delinquency more consistent
with the conventional understanding of the concept was the fact that only
monthly aggregates of the transactions of individual loans were available
at the time of structuring the transaction. Taking into account data limi-
tations, we defined delinquent loans as those loans that failed to pay their
aggregate monthly installments in the immediately preceding month.
The definition was stricter than BRAC’s own in that many BRAC
“current” loans were identified as delinquent. But on the other hand, it was
less strict than BRAC’s in that a loan that was beyond its one-year matu-
rity period but was making its payments on time in the immediately pre-
ceding month would not be identified as delinquent.
Table 3.3 shows the rate of delinquencies (defined as the principal out-
standing of delinquent loans over the principal outstanding of BRAC’s
total current portfolio) for the period January 2005 to April 2006. It also
shows the cash flow impact of delinquencies in this period. It must be noted
that Bangladesh did not suffer from any natural disasters in this period. But
even by our arguably stricter definition of delinquencies, the rate of delin-
quencies has never been greater than 8.25 percent.
60 Microfinance
Table 3.3 Delinquency rate and cash impact of delinquencies, January

2005 to April 2006
Month Year Delinquency Rate (%) Cash Impact (%)

1 2005 8.14 7.69
2 2005 7.48 7.33
3 2005 8.25 7.49
4 2005 7.81 7.35
5 2005 6.62 6.49
6 2005 6.14 5.53
7 2005 6.42 6.09
8 2005 5.93 5.98
9 2005 5.47 5.88
10 2005 6.51 6.62
11 2005 4.77 5.72
12 2005 4.24 4.57
1 2006 4.83 5.18
2 2006 4.45 4.83
3 2006 4.51 4.77
4 2006 4.89 4.88
Average 6.03 6.02
Source: BRAC data.
Analysis of risk
Risk variables identified In the available data, our analysis identified the
location of the borrower and the type of activity of the loan as relevant
risk variables. Borrowers in peri-urban and urban areas, for example, were
less likely to be delinquent than borrowers in remote parts of the country.
We also identified the age of the loan as a relevant risk variable. Loans
that were six to eight months old were more likely to be delinquent than
newer or older loans. We also found that loan size did have an impact
upon delinquencies. The larger loans given in the Progati program to small
enterprises almost never missed payments. Delinquencies were much more
frequent for the smallest loans in the Dabi program.
Prepayments Our analysis also revealed the existence of risks to the struc-
ture that had not been anticipated in the beginning of the structuring process.
We noticed that a significant number of loans prepay before their maturity
date. Table 3.4 shows the cash impact of these prepayments between January
2005 and April 2006. It is worth noting that in some months, the positive cash
impact of prepayments is comparable to the negative impact of delinquen-
Table 3.4 Prepayment rate and cash impact of prepayments, January 2005
to April 2006
Month Year Prepayment Rate (%) Cash Impact (%)

1 2005 7.73 4.22
2 2005 11.14 5.04
3 2005 5.84 2.81
4 2005 5.63 2.95
5 2005 8.39 4.36
6 2005 3.26 2.74
7 2005 2.69 8.83
8 2005 2.83 8.81
9 2005 5.17 10.82
10 2005 4.82 19.02
11 2005 10.49 21.17
12 2005 7.65 11.99
1 2006 4.58 10.11
2 2006 4.99 11.86
3 2006 5.07 10.01
4 2006 2.70 9.12
Average 5.81 8.99
Source: BRAC data.
cies. Our structure would need to take into account the risk of prepayments
not leaving enough cash flow for bond payments in future months.
PERFORMANCE
At the time of the writing of this chapter, 11 (of 12) payment dates on Tranche
1 of the transaction had passed, and five for Tranche 2. Over 99 percent of
Tranche 1 has already been paid down, and 60 percent of Tranche 2. We are
now in a position to look closely at the efficacy of the structure.
Delinquency Rates
Bangladesh has not witnessed any natural calamities in the last few months.
However, there has been some political turmoil over the parliamentary elec-
tions, which were cancelled on 11 January 2007 with the declaration of a
State of Emergency caretaker government.
Table 3.5 looks at the delinquency rate of loans in the securitized pool.
(Delinquency rate for a particular month is defined as the principal
62 Microfinance
Table 3.5 Delinquency rate for Tranche 1 and Tranche 2 securitized pool,
August 2006–June 2007
Month Tranche 1 (%) Tranche 2 (%)

August 2.60
September 2.80
October 8.20
November 6.10
December 4.00
January 5.80
February 6.90 2.90
March 14.00 5.30
April 18.00 7.30
May 15.00 6.40
June 16.00 7.90
Average 9.04 5.96
Source: BRAC Data.
outstanding of loans that missed a single payment in the previous month

divided by the principal outstanding of the securitized pool.) Delinquencies,
in the politically volatile period between November 2006 and January 2007
never rose above 6.10 percent. The relatively high delinquency rate in October
2006 (just after the flood season) is comparable to the rate from previous years.
Delinquencies for Tranche 1 hovered above 15 percent in March 2007. This
was because of the aging of the pool. A jump in the delinquency rate was
expected, as older loans have a much higher probability of being delinquent.
Bond Performance
Tables 3.6a and 3.6b show the excess cash flow in the first two tranches
returned to BRAC as residual beneficiary, and the future SPV liability
covered by the amounts collectible from the securitized pool and cash
trapped in the prepayment accounts.
It is clear that the SPV is awash in liquidity. This was expected, and was
one of the reasons that the credit rating report cited the AAA rating for the
bond. For both tranches, at least 150 percent of the SPV’s liability in future
months is covered by the underlying securitized receivables and the cash
trapped in the prepayment account without taking into account the US$2.5
million deposited in the debt service reserve account.
But what would have happened had the credit enhancements not been put
in place? Tables 3.7a, 3.7b, 3.8a and 3.8b examine the impact of removing
Table 3.6a Tranche 1 excess cash flow and SPV liability covered by pool
and trapped prepayments, September 2006–July 2007
Payment Year % Excess % Future SPV

Month Cash Flow Liability Covered
9 2006 99 157
10 2006 87 157
11 2006 62 166
12 2006 79 172
1 2007 56 184
2 2007 72 199
3 2007 78 222
4 2007 96 286
5 2007 125 413
6 2007 174 762
7 2007 339 1823
Note: DSRA (debt service reserve account) amount not included in any calculation.
Source: BRAC data.
Table 3.6b Tranche 2 excess cash flow and SPV liability covered by pool
and trapped prepayments, March 2007–July 2007
Payment Year % Excess % Future SPV

Month Cash Flow Liability Covered
3 2007 70 161
4 2007 84 171
5 2007 87 178
6 2007 84 190
7 2007 76 209
Note: DSRA amount not included in any calculation.
Source: BRAC data.
the different credit enhancements—over-collateralization, the trapping and

controlled release of prepayments, and the substitution and replacement of
delinquent loans.
Removing over-collateralization would have resulted in the SPV being
unable to cover the bond obligations beyond the first two months for both
tranches. The securitized pool barely covers future SPV liabilities in the first
Table 3.7a Performance of different credit enhancements for Tranche 1: % excess cash flow of SPV liability returning
to residual beneficiary under different scenarios
Payment Year All Credit Over-collat., Over-collat., Over-collat., No Over-collat.,

Month Enhancements No Cr. w/Prepay. w/Subst. w/No
Enhance. Trap. Cr. Enhance.
9 2006 99 102 87 89 35
10 2006 87 83 64 70 21
11 2006 62 40 24 44 ⫺7
12 2006 79 49 30 54 ⫺1
64
1 2007 56 49 16 30 ⫺1
2 2007 72 41 18 37 ⫺6
3 2007 78 37 12 37 ⫺9
4 2007 96 24 ⫺2 44 ⫺17
5 2007 125 13 ⫺18 54 ⫺25
6 2007 174 11 ⫺23 71 ⫺26
7 2007 339 26 ⫺14 154 ⫺16
Note: Calculations do not include DSRA amount of US$2.5 million.
Source: BRAC data.

Table 3.7b Performance of different credit enhancements for Tranche 2: % excess cash flow of SPV liability returning
to residual beneficiary under different scenarios

3 2007 70 75 61 65 17
65
4 2007 84 64 48 73 9
5 2007 87 47 29 70 ⫺2
6 2007 84 32 15 63 ⫺12
7 2007 76 16 ⫺1 52 ⫺23
Source: BRAC data.

Table 3.8a Performance of different credit enhancements for Tranche 1: % of future SPV liability for Tranche 1 covered
by amount collectible from securitized pool and cash trapped in prepayment accounts

9 2006 157 152 155 156 101
10 2006 157 147 153 154 98
11 2006 166 148 158 161 99
12 2006 172 148 165 165 99
66
1 2007 184 148 181 173 98
2 2007 199 150 202 182 100
3 2007 222 156 243 199 104
4 2007 286 152 306 251 101
5 2007 413 185 501 353 123
6 2007 762 266 1092 633 177
7 2007 1823 566 3320 1517 378
Source: BRAC data.

Table 3.8b Performance of different credit enhancements for Tranche 2: % of future SPV liability for Tranche 2 covered
by amount collectible from securitized pool and cash trapped in prepayment accounts

3 2007 161 155 157 159 103
67
4 2007 171 137 142 167 91
5 2007 178 134 144 172 89
6 2007 190 126 142 182 84
7 2007 209 127 152 198 84
Source: BRAC data.

68 Microfinance
few months. The trapping and controlled release of prepayments would

have allowed the securitized pool and trapped cash to cover future SPV liab-
ilities more comfortably, particularly after the fifth month of the transac-
tion. Substitution and replacement of delinquent loans have a similar
impact. However, the SPV still would not have been able to cover its liabili-
ties between the second and sixth month.
With over-collateralization but no other credit enhancements, Tranche 1
would have had some excess cash flow in all months, and the SPV would also
have been able to cover at least around 150 percent of its future liabilities in
any month. While the excess cash flow would have fallen to 15 percent in the
last few months of Tranche 1, the structure could easily have been adjusted to
accelerate payments in the earlier months where excess cash flow was higher.
With over-collateralization and substitution (but no trapping of prepay-
ments), there would have been a slight decrease in excess liquidity, which
would still hover around 30 percent in the fifth to seventh months of the
transaction.
Over-collateralization without substitution but with the trapping and
controlled release of prepayments would have meant less SPV liquidity
throughout the life of Tranche 1. In fact, the structure with the current
amortization schedule would have been unable to pay its obligations in later
months. However, the future liability covered by the pool and the trapped
prepayments would increase significantly in later months. The trapped pre-
payments could be used to extend the maturity of the bond to beyond 12
months. A viable structure could have been created with accelerated pay-
ments in the early months, lower months in the later months (where cash
flows would have been negative) and the release of prepayments to pay
bond obligations beyond 12 months.
The Tranche 2 experience is consistent with that of Tranche 1.The trans-
action till date has not required the replenishment of the pool with new col-
lateral, as excess cash flow has never dipped below 40 percent. We have thus
not analyzed the impact of replenishment upon the robustness of the
structure.
Clearly, much of the robustness of the structure can be attributed to
over-collateralization. The process of trapping prepayments and releasing
them upon maturity adds to the comfort that the SPV shall be able to cover
future obligations. However, even without the trapping of prepayments, at
least 150 percent of future obligations are covered by the receivables from
over-collateralized pool at any point in time for Tranche 1. Substitution has
a larger impact on SPV liquidity than the trapping of prepayments, but is
not crucial for a viable structure. These results suggest under normal cir-
cumstances, over-collateralization may be enough as a credit enhancement.
LOOKING FORWARD
Servicer Risk
We have discussed many of the informational and logistical risks that were
identified and addressed by the structure. However, a major risk noted by the
investors was servicer risk. In the event of BRAC ceasing to exist, it would take
a few months for another institution to step in and restart collections from
BRAC’s borrowers.
A few factors mitigate some of the servicer risk. PriceWaterhouseCoopers
(PWC) ran a systems analysis of BRAC’s MIS. PWC noted that BRAC’s
systems were quite robust. In particular, they noted the presence of data
back-ups and system redundancies at both the head office and area office
levels. They also independently checked the accuracy of the data. Further,
the BRAC database includes the name and location of the individual bor-
rower. Thus, in case the servicer needs to be replaced, the new servicer will be
able to identify and locate the borrowers whose loan has been securitized.
Yet, it is unavoidable that a new servicer will likely take some time to send
its own collection officers. Additionally, while all funds collected from the
securitized pool can be identified in the current BRAC system, the 42-day
time lag for information reaching the head office means that there is always
some commingling of funds in BRAC’s accounts. In the event of BRAC
going bankrupt, there is some risk of not being able to retrieve all the col-
lected funds for the immediately preceding month. However, the presence
of a DSRA fund may mitigate some of these risks.
Moving Beyond the BRAC Micro Credit Securitization Series I
We have already discussed how information on borrower characteristics and

households was limited. This constrained the risk analyses that could be done.
Further, historical data were unavailable. In future transactions, more will be
known about the risk profile of BRAC’s borrowers from its existing databases.
But, as noted, BRAC is not tapping into rich sources of information that
are readily available to it. BRAC’s health care and education programs
collect detailed information on BRAC’s group members and their house-
holds. However, currently these databases cannot be linked to the micro-
credit database. Further, BRAC also collects information on the savings of
the borrowers. These data were unavailable to us, and can yield valuable
information about the risk profiles of BRAC’s borrowers.
As more is learnt about micro-credit borrowers, investors also have to be
more flexible in their approach to conditions they place on future securiti-
zations. The 50 percent over-collateralization, for example, was dictated by
70 Microfinance
the investors at the start of the transaction without much analysis. The 40
percent excess cash flow requirement and DSRA amount were also a func-
tion of some investors’ (and regulators’) risk aversion and the need to
guarantee an AAA rating in this first-of-a-kind transaction. In future
transactions, investors should allow analyses of the risk profile of the bor-
rowers to guide optimal structures. Such analyses may also create the pos-
sibility of tiering risks to meet the appetites of different investors.
The sharing of information about borrowers among different micro-
finance institutions may also help these institutions better understand the
risk profiles of their own borrowers. It may also allow smaller institutions
to pool their portfolios to achieve the necessary scale for accessing funds
through securitizations. Sharing of information—particularly in a country
like Bangladesh with no national ID cards, let alone credit rating reports—
may help mitigate residual servicer risk.
We have noted how BRAC’s own MIS in the area and head offices is quite
robust, particularly given the difficult operational conditions. But the 42-day
lag between actual collections and information reaching the head office is
problematic. BRAC has made tremendous strides in getting the information
to the head office quicker. But more frequent updates would reduce the risk
of investors in future transactions, while also allowing BRAC to improve its
own cash flow management. BRAC is considering the possibility of linking
a few regional offices to the head office through the Internet, and transfer-
ring information to these regional offices from the area office weekly. It may
also be possible to leverage mobile technology for live updates on collections
and disbursements.
During the course of the transaction, the very fact that BRAC had to expose
and explain its processes and systems to investors, auditors, and a credit rating
agency helped advance transparency and accountability at BRAC. Over the
course of the transaction, issues spotted with BRAC’s systems or data were
rectified as far as possible. BRAC has also made tremendous strides in the last
year or so in strengthening its reporting systems. We predicted at the start of
this transaction in 2004 that BRAC would mature as a result of this securiti-
zation. We were pleasantly surprised when this prediction was confirmed.
A securitization such as this is really a learning process for the origin-
ator, investors, and regulators. It contains lessons for future transactions
at many different levels, from political economy considerations to risks
identified. The structure that was created for BRAC Micro Credit
Securitization Series I incorporated many of these lessons. The structure
has proved to be robust in the first six months of Tranche 1. As the trans-
action moves forward into its new tranches, we hope to learn more lessons
about what works in the current structure for replications of this transac-
tion and the creation of new MCBS structures elsewhere.
4. Cell phones for delivering
micro-loans
Anand Shrivastav
I will give you a talisman. Whenever you are in doubt, or when the self becomes
too much with you, apply the following test. Recall the face of the poorest and
the weakest man whom you may have seen, and ask yourself, if the step you con-
template is going to be of any use to him. Will he gain anything by it? Will it
restore him to a control over his own life and destiny? In other words, will it lead
to Swaraj [freedom] for the hungry and spiritually starving millions? Then you
will find your doubts and your self melt away. (Mahatma Gandhi)1
INTRODUCTION AND BACKGROUND
Cutting-edge technology can provide a transformational role in delivering

financial, health, and educational services to the poor. This chapter focuses
on India and the potential for providing financial services to the poor
through mobile phone technology. Such a delivery has the potential to cut
significantly the transactions costs of accessing small loans and this can
lead to an improvement in the overall welfare of the poorer sections of
society who are currently unable to access financial services from the organ-
ized banking sector and credit markets. To provide the reader with a back-
drop, it is first useful to review some background data on India.
Demography and Economy2
India is a large country with an area of 3.3 million square kilometers. The
population of India has exceeded 1.1 billion, and is showing a year-on-
year growth rate of about 2 percent. Of immediate interest to the subject
matter addressed in this chapter is the fact that India has a fairly high
savings rate (2004–05) of 29 percent, and a growing literacy rate. Moreover,
there is a large pool of professionals, and the youth population is significant
as attested by the fact that more than 50 percent of the population is under
25 years old. An important fact that has been recognized by the policy-
makers is that there is significant unemployment, especially in the age group
20–24 in rural and urban India.3 The unemployment rate of educated
71
72 Microfinance
women in rural and urban India is very high as well. Increased availability
of credit and other financial services is likely to be an important factor in
alleviating unemployment, and in empowering women.
Mobile Telecom4
Mobile telecom has already reached where other sectors are yet to reach.
Telecom statistics and growth projections are spectacular: mobile sub-
scribers have a cumulative annual growth rate of 86 percent and were
expected to reach 143 million by November 2006. The Department of
Telecom—DOT—puts forward a target figure of 500 million mobile con-
nections and mobile access to every village over 1000 population by 2010.
Despite this remarkable growth, the mobile penetration is only 10 percent
and is one of the lowest in the world although, even with the 400
minutes/subscriber/month usage, which is close to the United States figure,
Indian telecom is well placed in the world markets.
Looking at the future, potential projections vary. However, even a con-
servative estimate sees 300 million subscribers by 2010. Considering that
the mobile penetration is lower than many other comparable economies,
the growth rate of mobile phones in India is likely to remain very high for
the foreseeable future.
Microfinance5
The RBI (Reserve Bank of India) Internal Group’s “Report on Micro

Finance” has noted that the outreach of the Indian banking system has
seen rapid growth in rural areas. Forty-eight percent of the total branches
of the scheduled commercial banks (SCBs) and regional rural banks
(RRBs) cater to the rural areas (32 303 branches translates to a population
of about 23 000 people per branch). Of these, 31 percent (136.7 million) of
deposit accounts and 43 percent (25.50 million) of borrower accounts are
in the rural areas. This expansion of the organized financial infrastructure
has reduced the dependence of the rural population on the unorganized
moneylending sector from 68.3 percent in 1971 to 36 percent in 1991.
In spite of this growth, there continues to be wide gaps in the availabil-
ity of banking services in the rural areas as the SCBs cover only 18.4 percent
of the rural population through savings/deposit accounts and even a lower
percentage of 17.2 percent of the rural households by way of loan accounts.
Though the primary agriculture credit societies (PACS), with about 100 000
outlets, have a deep and wide presence in rural India, their impact in terms
of extension of deposit and credit products has not only been minimal but
concentrated in a few states only.
Cell phones for delivering micro-loans 73
The decline in productivity of the rural branches of the commercial

banks, fragility of the cooperative credit structure, and weakness of RRBs
witnessed since the early 1990s has further accentuated the problem of
inaccessibility of banking services for a large part of the rural population.
Furthermore, as the banking sector has shown, a propensity towards the
larger-size accounts has meant that the number of loan accounts of small
borrowers with a credit limit range of less than Rs 25 000 decreased from
58.8 million in 1991 to 36.9 million in 2003.
A vast majority of the rural population remains unintegrated with the
organized banking sector. On analyzing the supply side, we observe some
of the reasons for this, such as: (1) people are unbankable in the evalu-
ation/perception of bankers, (2) the loan amount is too small to invite the
attention of the bankers, (3) the person is bankable on a credit appraisal
approach but distances are too far for servicing and supporting the
accounts, and expanding branch network is not feasible and viable,
(4) there are high transactions costs, particularly in dealing with a large
number of small accounts, (5) there is a lack of collateral security, (6) an
inability to evaluate and monitor cash flow cycles and repayment capaci-
ties due to information asymmetry, lack of data base, and absence of
credit history of people with small means, (7) there are human resources-
related constraints both in terms of inadequacy of personnel and lack of
proper orientation/expertise, (8) there is an adverse security situation pre-
vailing in some parts of rural India, (9) a lack of banking habits and
credit culture, (10) information-shadow geographical areas (geographical
areas where information on identity and residential proof of persons is
yet to be completed, which makes know your customer (KYC) compli-
ance difficult for banks; hence the population in these areas remains
unbanked), and (11) an inadequacy of extension services, which are
crucial to improving the production efficiency of the farmers, leading to
better loan repayments.
Similarly on the demand side, there are several reasons for the rural poor
remaining excluded from the formal banking sector, such as: (1) high trans-
actions costs at the client level due to expenses such as travel costs, wage
losses, incidental expenses, (2) lack of awareness, (3) lack of social capital,
(4) non-availability of ideal products, (5) very small volumes/size of trans-
actions, which are not encouraged by formal banking institutions,
(6) hassles related to understanding documentation and procedures in the
formal system, (7) easy availability of timely and doorstep services from
money lenders/informal sources and (8) prior experience of rejection by
indifference of the formal banking system.
Under the microfinance program, loans are extended to the self-help
groups (SHGs) who pool a part of their income into a common fund from
74 Microfinance
which they can borrow. The members of the group decide on the minimum
amount of deposit, which ranges from Rs 20–100 per month depending
upon the size of the group. The group funds are deposited with a
microfinance institution (MFI) against which they usually lend and the
deposits are usually placed with a bank by the MFI. The group funds are
the way “micro-savings” are enforced, though it may seem like collateral.
The loan ticket sizes are usually Rs 2000–15 000. Although loan repayment
is a joint liability of the group, in reality individual liability is emphasized.
Maintaining group reputation leads to the application of tremendous peer
pressure.6
In India and other Asian countries the majority of SHGs typically
consist of women because, in these countries, self-employment through
microfinance was perceived as a powerful tool for the emancipation of
women. A World Bank report (2001—see note 6) observes that gender
equality is a necessary condition for economic development. It reports that
societies that discriminate on the basis of gender are in greater poverty, have
slower economic growth, weaker governance, and lower living standards.
And the results are encouraging. Loans obtained from MFIs are utilized in
agriculture and small businesses. Independent incomes and modest savings
have made women self-confident and helped them to fight poverty and
exploitation.
This can be seen from a statement by a woman beneficiary: “Previously
we had to cringe before our husbands to ask for one rupee. We do not have
to wear tattered sarees anymore and, today, we have the confidence to come
and talk to you without seeking permission from our husbands” (as told to
the author of the UNPAN’s Field Survey [note 6]).
Suvidha7 proposes to play a significant role in microfinance using the
SWIFT mobile transaction platform and providing Beam (see “The
Product” below) services, leveraging its distribution network and customer
profiles. Banks have the opportunity to partner with Suvidha to supplement
their extension banking services too. Likewise, MFIs, RRBs, SCBs, NBFCs
(non-banking finance companies), and banks can also partner Suvidha to
extend their product deliveries and manage both inbound as well as out-
bound payments through the network of Beam Mobile Entrepreneurs.
Payment System Inefficiencies
The GSM Association (GSMA) launched a pilot program in January 2006,

aimed at tapping the ubiquity and ease-of-use of mobile technology to
enable the world’s 200 million international migrant workers to send remit-
tances easily and securely to their dependants, many of whom did not have
bank accounts.8 By exploiting the extensive reach of the mobile networks,
Table 4.1 Payment realization for various payment modes in India
Nature of Financial Location of Time to Complete

Transaction Transaction Transaction
Cheques Local (same bank) Three days
Outstation 10–30 days plus seven
days by post.
Demand draft Local (same bank) One to two days, plus seven
days by post
Outstation Three days, plus seven days
(other bank) by post
Electronic
Credit card Seven days (merchant
payments)
Debit card Seven days (merchant
payments)
Electronic clearing system Four days (limited
coverage)
Postal money order 10–45 days
Source: Data collected by author.
the program complemented existing local remittances channels and made

transferring money internationally significantly more affordable.
In India, the situation of payment realization for different modes is as
shown in Table 4.1.
It can be seen from Table 4.1 that the present payment systems in India
are still quite inefficient and can contribute to lower than optimal rates of
economic growth.
GLOBAL TRENDS
As per a report by IFC Washington-GSMA, the advantage of developing a

market for micro-payments (also referred to as m-commerce), is that it
continues to drive the economic system toward a cashless transaction envi-
ronment.9 Elimination or minimization of physical cash has many advan-
tages, including less opportunity for fraudulent or criminal activity,
reduction of cash handling costs, and, for the user, less reliance on having the
right amount of cash when needed. It also allows the value of money to be
better utilized. Cash held outside the banking system is not available for
short-term investment so that the time-value of the cash asset is lost.
76 Microfinance
In the more affluent economies, there is already a good infrastructure for

a cashless environment, with most people having bank accounts and an
array of both debit and credit cards. Nevertheless there is an underlying
need for cash for minor purchases and there is little incentive to eliminate
cash entirely. These economies can manage quite well and there is no
specific interest group that feels sufficiently under pressure to develop
systems aimed at eliminating cash from the environment. Systems that have
been developed in such markets are often expensive and hence not particu-
larly attractive to the customer.
In the developing economies however, there is a very large “underclass”
that is totally reliant on cash for all their day-to-day expenses. Moreover,
this underclass makes no use of the banking sector and so is “invisible” in
terms of its cash value. At the same time, the need for cash forces the
providers of goods and services in these markets to have adequate cash-
handling facilities and this comes at some cost. In these cases, the com-
mercial organizations have much more to gain by addressing the problem
of cash transactions. Not only is the risk associated with cash holdings
much greater, but the time-value of the cash being held outside the banking
sector is entirely lost. Furthermore, the population in this category is lost,
that is, unseen by the banking sector. For these reasons, there is likely to be
more incentive in developing economies to move the population at large
away from cash, than exists in developed economies. That being so, a solu-
tion that meets the needs of developing economies will also have extensive
application in the developed economies. This arises because the solution
must be accompanied by very low costs as, if it were otherwise, the solution
would have no appeal in those developing economies. The resulting low-
cost solutions can then be applied in the developed economies, resulting in
further efficiency gains.
Further, according to the IFC report, the most successful micro-payment
applications are to be found in the Philippines, with over 3.5 million
m-commerce users on two mobile networks. The key success factors for that
market included the ability to load prepaid airtime credits as well as the
ability to transfer both cash and airtime credits between customers.
Coupled with these were the low values set by the operator for such prepaid
top-ups or credit transfers. Typical top-ups of 47–57 (US) cents were
allowed by the networks (equivalent to around four to five minutes of calls)
while transfers between customers of both cash and airtime credits were
permitted as low as 4 cents.
The target market surveyed by IFC is attuned to “sachet purchasing” or
the practice of purchasing goods in very small quantities packed in sachets.
This phenomenon is known to be common in other developing markets
where the populace rely on cash for all trading and can afford to buy
provisions for just a few days’ consumption. This market does not exhibit
bulk purchase tendencies and m-commerce that involves a significant cash
deposit or payment will be unlikely to find any significant uptake from the
target market.
While the application of m-commerce to developing markets was not
constrained to the Philippines, African (South Africa and Kenya) market
developments seemed to reflect the Filipino views, indicating that the target
markets in these geographically diverse areas were very similar in their use
of cash and their expectations.
The range of features available in each market showed significant uni-
formity as to be expected if the target markets were similar. With minor
variations, the features of all systems included:
● provision for cash deposits and withdrawals;

● the ability for third parties to make deposits into a user account
(employer, family member, or MFI);
● the ability to make retail purchases at selected outlets;
● over-the-air prepaid top-ups using the cash already in the account;
● the ability to transfer cash between users’ accounts;
● the ability to transfer airtime credits between users;
● provision for bill payments.
These features could be used for microfinance applications involving both

loan repayments as well as loan advances, and this area in particular is
being exploited in the Kenya trials and in Philippines services in conjunc-
tion with the Rural Bankers Association.
Apart from the use of the services by MFIs all services studied by IFC
operated on a debit account basis, that is, the account could only be oper-
ated in credit. As a result, bad debt is not an issue other than loss caused
by fraudulent activity. No operator indicated any serious concerns in this
area and provided the overall system security was ensured the possibility
for fraud could be managed. In that regard, most of the systems studied
involved a bank with normal banking systems in place as this arrangement
results in the fraud issue being restricted to the bank’s area of involvement
for which it will be well-equipped.
The possibility of money laundering was considered by all service
providers and it was noted that in all jurisdictions the banking regulatory
authority had established appropriate policies governing the activities
of the banks. These policies included monitoring transaction levels
and frequency, looking for transaction patterns and stipulating both
maximum account balances and daily transaction levels. While it is poss-
ible for a network operator to take almost full responsibility for the entire
78 Microfinance
Table 4.2 Increase in SMS traffic in Asia Pacific, 2004–10
SMS 2004 2005 2006 2007 2008 2009 2010

Traffic (billions) 434.1 540.1 672.8 802.4 935.9 1072.1 1212.7
Source: Portio Research, accessed at

www.adnetasia.com/news/communications/0,39044192,39252956,00.
micro-payment service, only one service in the Philippines was operating in

that manner. Even then, the actual cash float generated was held in one of
the country’s regular banks. All other cases studied involved cooperative
arrangements between banks and networks. In view of the regulatory issues
surrounding the banking industry, this method of operation is more likely
to appeal to intending service providers, given that the banks can bring
additional advantages including the availability of debit cards through
issuers such as MasterCard.
While there were no quantifiable figures available on system costs,
various estimates placed a likely cost in the range of US$5 million to US$10
million with an expectation that an m-commerce system could be profitable
with as few as 25 000 users connected but that would depend on the overall
investment and service operating costs. Various estimates placed the trans-
action level at around two per customer per day and average transaction
values at between US$15 and US$30 per customer and airtime top-ups of
around US$4 per time. These figures can only be regarded as indicative of
the type of activity that may be encountered.
According to a study by Portio Research, SMS (short message service)
traffic in the Asia Pacific (AP) region is expected to increase from 434
billion messages in 2004 to over 1.2 trillion by 2010.10 (See Table 4.2.)
One of the key growth drivers is attributed to higher future mobile
phone penetration in India and China. The report projected that the com-
bined markets of India and China would account for over 800 billion
SMS messages per annum by 2010. The study predicts that India, with a
low mobile penetration rate has “massive potential.” SMS in India is
expected to grow from 12.3 billion messages in 2004, to 180 billion in
2010.
In summary, the situation provides an excellent and growing opportunity
to Suvidha and its SWIFT technology platform, which uses SMS to
perform both inbound as well as outbound transactions. Suvidha’s network
of Beam Mobile Entrepreneur women become self-employed by providing
the front end services at the localities where they live, and thus gain better
control over their lives.
Table 4.3 Domestic transaction companies operating in India
Company Website
Paymate www.paymate.co.in/web/Default.aspx
JiGrahak www.jigrahak.com/site
MChek www.mchek.com/
ITZ Cash www.itzcash.com/
Done Card https://www.donecard.com/index-1.aspx
Wallet 365 www.timesofmoney.com/tomsvc/jsp/home.jsp
Fino www.fino.co.in/
A Little World www.alittleworld.com/
Source: Gathered from market intelligence.
THE INDUSTRY
India currently has around eight domestic companies operating in the

transaction space as shown in Table 4.3.
Paymate is SMS-based but exclusively for use by customers having
accounts with banks, for example, Citibank. JiGrahak requires download
of software to a mobile handset and is thus limited to certain types of hand-
sets. MChek works on GSM technology, as it is USSD (unstructured sup-
plementary source data) technology-based solution and provides services
to the subscribers of Airtel and some banks. ITZ Cash and Done Card are
prepaid card-based solutions that work on the Internet. Wallet 365 is also
an Internet-based service for banked customers. Fino is a closed-user pro-
prietary technology service provider for MFIs. Last but not least, the tech-
nology focus of A Little World is on biometrics-based ID, RFID (radio
frequency identification) smart cards (Java, PKI) and NFC (near field com-
munication) mobile phones as acceptance and enabling devices (with mer-
chants, field forces of MFIs, and at cashless ATMs).
While ITZ Cash, Done Card, Paymate, and JiGrahak market themselves
as m-commerce, their services can only be used if the subscriber has Internet
access—microfinance is not their focus. Further, none are interoperable or
neutral as they are exclusively tied either to a specific bank (Paymate, MChek,
A little World) or telecom (MChek, JiGrahak) or can be used by special types
of phones (A Little World, JiGrahak, MChek via GSM), thus excluding a
significant number of CDMA (code division multiple access) subscribers.
Except for A Little World, none are focused towards microfinance.
Fino, on the other hand, acts only as an application service provider
(ASP) for microfinance institutions requiring core banking application and
80 Microfinance
point-of-sale devices. The parentage of Fino is ICICI Bank, hence possi-

bilities of interoperability with other banks will be a challenge.
Suvidha-Beam is focused on the micro-payments for the unbanked and
enabling microfinance offered by MFI/banks. Beam does not require the
Internet, or software to download, change of SIM, or require special equip-
ment. It is interoperable with subscribers of any telecom.
THE PRODUCT
Beam is an innovative and simple way of transacting money using mobile

phones. It takes advantage and plugs the inefficiencies in the payment
systems of the economy. It enables subscribers to register and use a host of
other services, anytime, anywhere, using short message service (SMS).
The product has two parts. A robust, future-ready technology platform
called SWIFT is at the backend; and a stored value prepaid card that con-
sumers purchase for using services, called Beam.
Backend Technology
The backend SWIFT (subscriber wireless interaccount financial transac-

tion system) is a mobile commerce platform that took several years to
develop. It leverages the cumulative knowledge and experience gained by
Suvidha from product distribution and providing transaction management
services to banks as well as telecom services. SWIFT is a sophisticated,
robust, secure, and scalable application having disaster management and
business continuity system too. It lets subscribers use the Beam services via
SMS, IVRS (interactive voice response system) or the Internet.
Services
Beam as a service allows mobile phone subscribers to send money, give
gifts, pay each other, make purchases from merchants besides a host of
other services—take credit, make deposits, obtain insurance, make invest-
ments, all using their mobile phone.
Services can be accessed as soon as a mobile phone customer registers
with Beam. This can be done by sending a simple SMS message to Beam
and assigning him or herself a secure personal identification number
(SPIN). The subscriber’s Beam account is established automatically by
SWIFT and the subscriber receives an SMS to this effect within a few
seconds.
Beam prepaid cards can be purchased from any retailer, Beam Merchant,
Beam Express (shops) franchisee or Beam Mobile Entrepreneurs (individ-
uals). Additionally the Beam prepaid cards can also be purchased from
Suvidha’s alternative channels comprising SCBs, cooperative banks, RRBs,
NBFCs, MFIs, and India Post.
Subscribers not having a bank account can purchase Beam prepaid cards
to top up their account and perform a variety of transactions. Money can
be gifted via Beam to another subscriber. A Beam Merchant can be paid by
a subscriber using Beam. Similarly, refund of the residual amount in the
subscriber’s Beam account can also be taken from any Beam Mobile
Entrepreneur or Express franchisee.
Besides micro-payment services, Beam Mobile Entrepreneurs can also
extend microfinance, micro-insurance, micro-investment as well as internat-
ional money transfer services of Suvidha partners to the Beam subscribers.
Additionally, the Beam Mobile Entrepreneurs can act as service delivery
vehicles and extend microfinance, micro-insurance, micro-investment, as
well as international money transfer services in his or her locality. These will
be to the customers located anywhere who may not have a mobile phone
and/or may not be registered with Beam but are clients of banks, coopera-
tive banks, RRBs, NBFCs, India Post, MFIs, SHGs, and Ladies’ Kitty
Clubs (LKCs).
Transaction Ecology
Figure 4.1 shows the human ecology of the various types of transactions.
Subscribers can be seen sending money, giving gifts, paying each other,
making purchases from member Beam Merchants and also taking refunds
from Beam Mobile Entrepreneurs using their mobile phones.
Similarly, Beam Mobile Entrepreneurs can be seen providing refund ser-
vices and also extending microfinance, micro-insurance, micro-investment,
and international money transfer services to Beam subscribers, as well as
to the customers of banks, SCBs, RRBs, NBFCs, MFIs, SHGs, India Post
and LKCs, who may have mobile phones and/or are not registered with
Beam.
THE FUTURE AND CHALLENGES
Suvidha is starting off with micro-payment services. As it moves forward,

it will use the flexibility and scalability of SWIFT technology platform,
leverage the distribution network and the profiles of Beam subscribers to
offer microfinance products, micro-insurance (life and general insu-
rance products like crop insurance and so on) of its partners. As the feet-
on-street, that is, Beam Mobile Entrepreneurs mature, it will offer the
82 Microfinance
Intl money transfer

Insurance
CREDIT - F2S Blank
Mutual fund
Member Mobile entrepreneur Mobile entrepreneur DEPOSIT - S2F
merchant franchisees franchisees INSURANCE - S2F
PAY - S2M
REFUND-S2F INVESTMENT - S2F
INTL MONEY XFR - F2S
Ladies-kitty Clubs
Self-help groups
Subscriber GIFT - S2S Subscriber
Note: OS2M = Subscriber to Merchant; S2F = Subscriber to Franchisee; S2S = Subscriber

to Subscriber; F2S = Franchisee to Subscriber.
Source: © Anand Shrivastav.
Figure 4.1 The ecology of Beam transactions
micro-investment products of partners. Suvidha will also offer internat-

ional money transfer services of partner money transfer organizations
(MTOs) through the Beam Mobile Entrepreneurs and Beam Express
franchisees. In addition to offering partner products, Suvidha will provide
micro-payments transaction management services to customers of banks,
SCBs, RRBs, NBFCs, microfinance, micro-insurance, micro-investment,
and MTO companies. It will move to other countries at an appropriate
stage.
Challenges
1. Regulation. The regulatory environment for payments, microfinance,

micro-insurance, micro-investment, and international money transfer
is still evolving and there are no clear guidelines.
2. Taxation. While service tax is well understood, VAT is administered by
individual states who are not clear on what the treatment should be
with regard to charging or not.
3. Anti-money-laundering (AML). Customers may not yet have been
issued the required documents with regard to anti-money-laundering.
4. Combating the financing of terrorism (CFT). Here again no clear dis-
semination of information has occurred at the enforcement level.
Micro-payments
Microfinance
Micro-insurance
Micro-investment
International money transfer
Source: © Anand Shrivastav.
Figure 4.2 Services to Beam subscribers
Nevertheless, Suvidha believes its services will not only improve the lives of
its customers and the Beam Mobile Entrepreneurs, but will enable them
have a greater control over their life and destiny, and in this manner make
its humble contribution to the economic prosperity of India.
NOTES
1. Mahatma Gandhi (1958), Last Phase, vol. II, p. 65. From a note left behind in 1948.
2. Census of India.
3. Government of India-MOSPI (2004), “Socio-economic dimension of unemployment in
India”; Government of India-MOSPI (2005) “Employment and unemployment situ-
ation in Cities and Towns in India 2004–05 Part 1”.
4. Sources include Telecom Regulatory Authority of India—TRAI June 2006;
Government of India Department of Telecommunication (2006), “Telecom Vision
2010” and Merrill Lynch (2006), “Global Mobile Matrix 2006”.
5. Sources include RBI (2005), “Internal Group Report of the RBI on issues relating to
rural credit and micro-finance 2005”; RBI Banking Statistics 2003; National Sample
Survey Organization (NSSO) (1999), “All India debt and investment survey 1991”;
National Federation of State Cooperative Banks.
6. Sources include UNPAN/Ghosh, R. (???) “Field Survey”; Sinha, F. and team (2003),
“Impact assessment of microfinance in India”, EDA Rural Systems Pvt Ltd, World Bank
(2001), “Engendering development through gender equality in rights-resources and voice”;
Sampark (2003), “Mid-term impact assessment study of CASHE project in Orissa”.
7. Suvidha (Sue – vee – dhaa) is a Sanskrit word that means convenience. Suvidha is a
mobile transaction service provider and is ISO 9001 certified. It was incorporated on 11
84 Microfinance
December 2002 and is headquartered in New Delhi. Its global transactions management
system (GTMS) is used by corporate banking for cash, tax management, and coopera-
tive payment management services via the Internet. Some of the banks using GTMS ser-
vices are HSBC, Deutsche Bank, ICICI Bank, HDFC Bank, and UTI Bank. Suvidha is
about to launch the Beam services in FY 2007–08, with SWIFT as its backend.
8. GSMA (2007), “Global money transfer uses pilot mobile to benefit migrant workers and
the unbanked”, press release, www.gsmworld.com/news/press_2007/press 07_14.shtml.
9. IFC Washington-GSMA-infoDev (2006), “Micro-payment systems”, infoDev report,
www.infodev.org.
10. ZDNet-Portio Research (2005), “SMS traffic to double in AP by 2010”, September,
www.zdnetasia.com.
5. How should governments regulate
microfinance?
Richard Rosenberg1
INTRODUCTION
Powerful new microfinance techniques are being developed that allow

formal financial services to be delivered to low-income clients who have
long had no access to such services. But the microfinance industry will not
reach its full potential unless many of its service providers can eventually
enter the arena of licensed, prudentially supervised financial intermedi-
aries. Regulations must eventually be crafted that allow this to happen.
Dozens of developing and transition country governments are now at
earlier or later stages of addressing this challenge.
Many different actors are pushing for regulatory adjustments, from
microfinance institutions themselves (MFIs), to international development
agencies, to government officials who want to democratize finance or
protect against perceived risks for the financial system (or perhaps clamp
down on annoying non-governmental organizations—NGOs). The inter-
ests and objectives of these actors diverge considerably. Thorny technical
and practical issues are involved. We do not have decades of experience
with regulated microfinance to guide us—most of the countries with
microfinance regulations have only a few years of experience with imple-
menting them. And country-specific circumstances loom large, so there can
be no standard model for microfinance regulation.
Nevertheless, among people working on these topics there are sur-
prisingly wide areas of consensus on some general principles that should
bear on regulatory design for microfinance. The author is confident
that most of the material in this chapter is consistent with the views of
most of the technical advisors who have multi-country experience and
who do not represent the interests of any particular combatant in the
fray.
The discussion will begin with an important definitional distinction
between “prudential” and “non-prudential” regulation. Non-prudential
regulation will be treated in the third section, prudential regulation in the
85
86 Microfinance
fourth, and the challenges of prudential supervision in the fifth. The sixth
section concludes.
PREAMBLE: PRUDENTIAL AND NON-PRUDENTIAL

REGULATION
Governments regulate the behavior of all businesses. Such regulation may
be aimed at protecting consumers, or employee safety, or the environment,
but it usually does not try to protect the financial health of the business—
that concern is generally left to the owners, at least where the owners are
private. But in almost every country in the world, banks are treated
differently. Governments impose an elaborate regime of “prudential” regu-
lation whose aim is to protect the solvency of banks. Various reasons are
advanced for this. The main one is that financial systems depend critically
on confidence, so that the failure of one bank can hurt many other banks
and provoke a systemic crisis, with dire effects for the economy at large.
Another reason is that banks are financed predominantly by money of
people other than the shareholders, which creates incentives for bank man-
agers to take imprudent risks: when the gamble succeeds the bank and its
shareholders capture the gain, but when the gamble fails, others—
especially depositors—may bear a large part of the loss. Finally, govern-
ments believe that small, unsophisticated depositors need protection
because they are in no position to appraise the riskiness of a bank on their
own. Examples of prudential requirements include capital adequacy rules
(how much of other people’s money a bank can use), restrictions on risky
uncollateralized lending, limits on insider lending, or requiring main-
tenance of reserves for loans that are likely to go bad.
“Non-prudential” regulation is an inelegant name for all the other
banking rules—the ones that don’t involve the government in assessing and
protecting the financial health of banks. Such rules are sometimes referred
to as “conduct of business” regulation. Examples include limits on interest
rates charged to borrowers, other consumer protection like truth-in-lending
laws, or anti-money-laundering rules that require screening and reporting
of customers. Securities regulations are another non-prudential example:
these rules usually require banks, just like other firms, to disclose all mate-
rial information about their business to potential investors. The rules have
been complied with when all the information about the bank’s business is
disclosed. Investors are left to fend for themselves when it comes to weigh-
ing their risks.
Implementing prudential regulation, where the government in some
sense is vouching for the financial soundness of each licensed bank taking
How should governments regulate microfinance? 87
deposits, tends to be much more complex, difficult, expensive, and intrusive

than implementing non-prudential regulation. Enforcing prudential regu-
lation always requires a specialized banking authority, whereas many non-
prudential regulations apply to non-deposit-taking firms as well, and might
not necessarily require a specialized banking supervisor to enforce them.
The reason for emphasizing this distinction is that in a number of coun-
tries, governments are applying burdensome prudential rules to non-
deposit-taking MFIs whose failure would cause neither loss of depositors’
funds nor material disruption of the national financial system. Prudential
regulation has high costs not only for the supervisory authority but also for
the supervised institution, which will eventually pass these costs along to its
customers. Prudential regulation of credit-only MFIs uses a cannon—a
very expensive cannon—where a rifle would be more than adequate in view
of the risks involved.
It is especially important to focus on the implications of regulation for the
administrative costs of MFIs. In the centuries-old effort to improve financial
access for poor and low-income people, the critical factor is cost, more than
the motivations of financial service providers. Major, long-lasting improve-
ments in access are usually associated with new ways to lower costs. Concern
for the poor has played an important part in the microfinance revolution of
the last three decades. But concern for the poor has been around for a long
time. The revolution became possible when Grameen Bank and other pion-
eers in Indonesia and Latin America discovered less costly ways to deliver
and collect tiny uncollateralized loans, and mobilize and manage small
savings. Some of the regulatory requirements discussed in this chapter have
significant cost implications for microfinance providers. Decisions about
such practices need to be handled carefully.
NON-PRUDENTIAL ISSUES
Usury Limits
Lending a million (US) dollars in 10 000 loans of $100 each entails admin-
istrative costs that are hugely greater than the cost of lending out the same
amount in one or two big loans. As a result, it is usually impossible to do
micro-lending on a financially sustainable basis without charging interest
rates that are very substantially higher than what banks charge to larger
borrowers. In 2005, the median annual interest rate collected by the hun-
dreds of MFIs reporting to the MIX Market (www.themix.org) was about
31 percent. Rates above 50 percent are not uncommon, and a few MFIs
charge more than 80 or even 100 percent. In most cases, these rates reflect
88 Microfinance
not high profits but high costs of micro-lending: the smaller the loan size,
the higher the administrative costs are for lending a given amount. But this
analysis of lending costs is fine print that is usually too small to show up on
the screens of politicians or the general public. Charging poor borrowers
30 percent when fat cats pay only 10 or 15 percent shocks most consciences.
Not all microfinance interest rates can be explained by the costs of lending.
In a recent well-publicized instance, the Mexican MFI Compartamos was
charging interest of about 100 percent a year and producing annual profits
that gave its shareholders more than a 50 percent return on equity. This cause
célèbre, despite being a highly exceptional case, has fanned the flames of a
growing backlash against high micro-credit interest rates, a backlash that has
already been underway in Latin America and the rest of the world for several
years now.2
Limits on interest rates can hurt rather than help low-income borrowers
if the interest cap is set too low for certain types of lending to be profitable:
providers will withdraw from the business and potential borrowers will lose
access to services. In theory, an interest rate cap would avoid this result if it
were set at just the right level. As a practical matter, however, finding the
right level is hard, not least of all because loan products, clienteles, and
costs vary. Moreover, it is politically difficult for governments to set inter-
est caps at appropriate levels: a reasonable interest rate for tiny, high-cost
micro-loans will inevitably seem exploitative to most people, because they
do not understand the reasons for the high rates.
Consumer Protection and Borrowers’ Rights
When governments are concerned about microfinance interest rates that

sound abusive, they are sometimes advised to avoid interest rate caps and
focus instead on other borrower protection issues such as truth-in-lending
(disclosure of the full cost of loans in a format that makes it easy to
compare rates offered by various lenders) or prohibition of certain unac-
ceptable lending and collection practices.
Most MFIs today do not quote their loan charges in the form of an
annualized effective interest rate that includes all costs. In some cases there
may be a legitimate concern that explicit quotation of rates this way would
lead to a political backlash, resulting in interest rate caps that would make
it impossible to continue serving their clients. In most cases, though, micro-
lenders’ opposition to truth-in-lending policies and requirements probably
stems mainly from normal, less noble motives. Loan-cost disclosure may
not be a panacea, however: there are some indications that low-income
clients have trouble understanding and using the information. This concern
has led to scattered efforts to provide financial education for consumers, but
most of these programs are still too young to be assessed, at least in devel-
oping and transition countries.
Consumer protection regimes may also include restrictions on the way
loans can be made and collected. Obvious examples would include pro-
hibiting the use of violence or other heavy intimidation to collect loans, but
other less dramatic practices may also be deemed abusive. In South Africa,
for instance, so-called “micro-credit” consists mainly of firms making
high-interest consumer loans to a clientele consisting mostly of salaried
formal-sector employees.3 Taking possession of a borrower’s ATM card or
requiring delivery of a post-dated check for the loan amount were common
practices, which were prohibited under a non-prudential regulatory regime
created by the Micro Finance Regulatory Council (MFRC), a government
agency lodged outside of the banking authority. MFRC rules were given
teeth by stipulating that any loans issued in violation of those rules would
be legally unenforceable.
The Bolivian microfinance sector suffered huge losses when profligate
Chilean-backed consumer lenders started marketing to unsalaried micro-
entrepreneurs, passing out loans that bore no relation to the borrowers’
repayment capacity. Many borrowers got in over their heads, and since a
large percentage of these were also borrowing from more responsible
MFIs, those sound MFIs were badly hurt by the ensuing wave of defaults,
not to mention borrowers who lost their credit rating. The Bolivian
Superintendency of Banks responded by requiring all uncollaterized
lending to include an assessment of repayment capacity.
A the time of this Bolivian crisis, the government’s credit reference
bureau was not working well, so it was hard for lenders to know whether a
potential borrower had loans outstanding from another source, or had a
history of repayment problems. After the crisis, all of the actors found
themselves considerably more enthusiastic about the credit bureau, and
unlicensed lending-only MFIs were allowed to participate for the first time.
As a general matter, credit reference bureaus are a powerful tool for extend-
ing credit access to previously excluded groups, because the bureaus
significantly lower the costs of appraising borrower creditworthiness, and
strengthen borrowers’ motivation to repay. Credit bureaus make it possible
to lend to customers who would have been unprofitable otherwise.
Other consumer protection measures include privacy protection and
accessible dispute-resolution systems.
AML/CFT Regulation4
The Financial Action Task Force (FATF) is an international body that rec-
ommends standards for national legislation on anti-money-laundering and
90 Microfinance
countering the financing of terrorism (AML/CFT). The FATF standards

do allow room for adjustment to fit individual country circumstances, but
developing and transition countries are often nervous of straying far from
the standards, because winding up on the list of non-complying nations can
have severe consequences. Among the FATF standards are know-your-
client rules (ascertaining and documenting the customers’ true identities
and addresses), heightened surveillance of transactions, preserving trans-
action records, and reporting suspicious transactions to national authori-
ties. Many banks complain loudly about the additional costs generated by
these requirements when dealing with their normal customers. In the world
of microfinance, where transactions and balances are much tinier, full
enforcement of regular AML/CFT standards would make it uneconomic
to serve large groups of customers. The additional administrative costs are
particularly problematic when transactions are so small, and compliance
can be impractical for some kinds of clients. For instance, it is challenging
to document identity and address for people who have no national identity
card, are illiterate, and have never seen any document that specifies where
they live.
In its early years FATF was dominated by people coming from a law-
enforcement perspective, who were not always instinctively sympathetic
with concerns about how FATF rules might exclude low-income clients
from services. More recently, this problem is getting more attention at inter-
national and local levels. Taking a risk-based approach to AML/CFT, it
would not seem that transactions of, say, $30 or loan or savings accounts
of $300 create substantial security risks. Governments should consider
softer requirements, or waiving them altogether, for accounts and balances
below defined limits. After this was done in South Africa, for instance,
banks were able to offer basic, no-frills transaction accounts that in a few
short years reached 1.6 million customers, most of whom would have
remained unbanked if the AML/CFT rules had not been relaxed.
PRUDENTIAL ISSUES
Whether, When, and How to Open Prudential Licensing Regimes for

Microfinance
Regulation as promotion
In more than a few countries, the microfinance sector consists mainly of
weak non-governmental organizations that provide lending only, as well as
credit unions or similar savings and loan cooperatives, few of which are
large and stable. When a government is confronted with this situation, and
wants to catalyze a large expansion of quality financial services for its

lower-income population, one plausible-sounding response is to develop a
new licensing window that allows institutions to become prudentially regu-
lated, and take deposits, without facing minimum capital requirements as
high as those required for a full banking license. (Deposit-taking is doubly
important: not only does it provide a large funding source for expansion of
lending, but it also gives the institution’s low-income clients access to a
savings service that is often even more valuable to them than credit.)
In such countries, this “build it and they will come” approach is based on
the hope that the special licensing window will attract new private sector
entrants to the business, or encourage weak existing MFIs to tighten up
their operations so as to meet the prudential standards for licensing. There
is considerable controversy over this approach. It is premised on the belief
that the binding constraint is absence of appropriate regulation, rather than
scarcity of competent retail operators. International experience to date has
been too limited to produce a general answer to the question; if anything,
it suggests that the answer will vary from country to country.
For instance, Tanzania spent a great deal of time, effort, and money on
the development of a well-conceived licensing regime, but years afterward
the results were disappointing. South Africa does not offer a micro-banking
license, but the government took steps a decade ago to make it possible for
microfinance institutions to offer small loans at relatively high interest rates.
Despite this change, South Africa still does not have many solid institutions
offering uncollateralized loans to unsalaried micro-entrepreneurs.
Pakistan has a huge unserved microfinance market, and it’s hard to find
many countries with as good a licensing regime for microfinance as the one
Pakistan enacted in 2001. But until very recently, institutions licensed
under this law contributed hardly at all to the growth of microfinance in the
country, and the overall financial condition of licensed and unlicensed
providers was worse than it had been when the law was passed. Within the
last year, however, things are looking brighter: some of the newly licensed,
privately owned microfinance banks are expanding aggressively, and the
overall financial performance of the sector is improving.
Regulation that follows, rather than leads, the market

In most of the countries with effective prudential regimes in place today for
microfinance, the regulation came after, not before, the development of a
critical mass of strong MFIs that were delivering loans on a sustainable basis.
Bolivia has the longest and most solid record of successful microfinance
regulation, and this experience is often held up as an example where a new
licensing window for microfinance was a powerful contributor to the success
of the industry. But the Bolivian licensing regime was put in place only after
92 Microfinance
the country already had a number of strong NGO MFIs who had shown they
could manage their lending business stably and profitably. BancoSol, the
leading MFI, did not use the microfinance licensing window—it got a regular
commercial banking license well before the specialized microfinance law was
passed. Most of the other MFIs who got licenses under the new law could
probably have raised the money for a banking license if the easier and
cheaper MFI license had not been available. In BancoSol’s first year or
two, there was a certain amount of “supervision by winking,” as the
Superintendent waived application of some prudential rules that didn’t fit
microfinance very well. But the time the microfinance law was passed and
new prudential norms formalized, the Superintendency already had experi-
ence from supervising BancoSol. It’s possible to argue that Bolivian
microfinance did not need a new specialized microfinance window to reach
its present vital state, and that a few adjustments to the country’s banking
law and regulations would have created the necessary regulatory space.
When countries design a new licensing window for microfinance on the
expectation that licenses will go mainly to existing NGO MFIs during the
early years, the regulators sometimes don’t pay enough attention to the con-
dition of those MFIs and their loan assets. In Zambia, for instance, the
foreign aid agencies of the United States and Sweden financed development
of a prudential regime in 1999 that would allow MFIs to take deposits. But
at that time, sources say, the country had few if any MFIs whose cost recov-
ery and loan collection would make them safe custodians of customers’
deposits. There may have been some expectation that donor-funded tech-
nical assistance would turn the MFIs into strong institutions, but it is hard
to find many examples of weakly managed MFIs that have been turned into
vibrant, stable MFIs by outside technical assistance. This is not to suggest
that such assistance is useless: MFIs that already have strong managers
make good use of such support, but technical assistance can seldom turn a
weak manager into a strong one. Political considerations prevented enact-
ment of the Zambian law at the time. A set of microfinance regulations was
finally put into force in 2006, but a review of the MIX Market database as
of the beginning of 2007 shows only a single sustainable Zambian MFI,
and that one had only 12 500 clients.5
Adjusting Prudential Norms to Fit Microfinance Products and Institutions
Some regulations common in traditional banking need to be altered to

accommodate microfinance. Whether microfinance is being developed
through specialized stand-alone deposit-taking MFIs, or as a product line
within retail banks or finance companies, the following norms will usually
need re-examination:
Minimum capital
The kind of investors who are willing and able to finance MFIs may not be
able to come up with the minimum capital required for a full banking
license, especially as minimum capital requirements trend upward around
the world. Setting a low minimum capital bar is often the central objective
of those pushing for new licensing regimes for microfinance.
When banking authorities set minimum capital, bank safety and sound-
ness may not be their primary concern. Rather, minimum capital is often
used as a rationing device to manage the number of separate institutions
that have to be supervised. The arguments for and against low minimum
capital for MFIs will be treated in the next section, which deals with super-
visory challenges.
Capital adequacy
Under the Basel Accords, the relationship between shareholders’ equity
and bank risk assets is the foundation of prudential regulation. Equity is
treated as a cushion that protects depositors and other creditors of the
bank: the more of its assets are funded by shareholders’ money, the higher
the losses the bank can sustain and still be able to repay its depositors.
There has been controversy about whether solvency (capital adequacy)
requirements should be tighter for specialized MFIs than for banks. If
we want a level playing field in the financial sector, should microfinance
be penalized with tougher solvency requirements that lower shareholder
profitability?
Several theoretical arguments point in the direction of higher equity-to-
risk-assets ratios for MFIs. In the first place, deposit-taking microfinance is
a new business in most countries, which supervisors—and some MFI man-
agers as well—do not have decades of experience with. Second, most MFI
loan assets are not collateralized. Normally, MFI portfolio quality is very
good, but if an MFI starts to have problems with loan delinquency, they
can balloon out of control much faster than would be normal with collat-
eralized loans. Third, administrative costs for MFIs are much higher than
for commercial banks. When a significant part of the MFI’s loans are not
being paid, the uncompensated administrative cost on those loans decapi-
talizes the MFI faster than would be the case with a normal bank. All of
these considerations suggest tighter solvency requirements, at least in the
early years.
Balanced against those theoretical arguments is the clear empirical fact
that licensed MFIs suffer fewer loan losses than commercial banks do.
There is also emerging evidence that licensed MFIs are more resilient than
commercial banks in times of financial or economic emergencies. In a
recent banking crisis in Bolivia, all the commercial banks went insolvent
94 Microfinance
and MFIs came through in good shape. During the financial meltdown in
Indonesia, repayment plummeted on the commercial loans of Bank Rakyat
Indonesia (BRI), while there was hardly a blip in the repayment of its
micro-loans. When times are uncertain, low-income people are especially
anxious to maintain their access to a credit facility, which can be a life-saver
if an unexpected shock hits. BRI’s micro-borrowers understand that the
only way to keep access to a future loan if and when they need it is to faith-
fully repay today’s loan.
Some microfinance is delivered through credit unions and other financial
cooperatives. Application of capital adequacy norms to these institutions
presents a particular issue with respect to the definition of capital. All credit
union members have to invest a minimum amount of “share capital” into
the institution. But unlike an equity investment in a bank, share capital can
usually be withdrawn whenever a member decides to leave the credit union.
From the vantage of institutional safety, such capital is not very satisfac-
tory: it is impermanent, and is most likely to be withdrawn at precisely the
point where it would be most needed—when the credit union gets in
trouble. Capital built up from retained earnings, sometimes called “institu-
tional capital,” is not subject to this problem. One approach to this issue is
to limit members’ rights to withdraw share capital if the credit union’s
capital adequacy falls to a dangerous level. A more conservative approach,
now recommended by the World Council of Credit Unions, is to require
credit unions to build up a certain level of institutional capital over a few
years, after which time capital adequacy is based solely on those retained
earnings.
Unsecured lending limits and loan-loss provisions

The experience in normal banking is that loans are more likely to default
when they are not backed by collateral or guarantees. Thus, banking regu-
lations often put tight limits on unsecured lending—for instance, capping
it at no more than 100 percent of the bank’s equity base. Such a limitation
would make a portfolio of uncollateralized micro-credit impossible, at least
in a specialized MFI. Some regulators have avoided the problem by treat-
ing group guarantees as collateral. But not all micro-lenders use a group
methodology, and group guarantees are less effective than is commonly
supposed. Many MFIs do not really enforce such guarantees, and loan
losses in MFIs that use such guarantees are not markedly lower than loan
losses in MFIs that do not.
A more straightforward approach is to recognize the empirical evidence.
Worldwide, with relatively few exceptions, uncollateralized loans in a
country’s licensed MFIs suffer less delinquency and default than collater-
alized loans in normal bank portfolios. The reasonable response to this
evidence is to put no collateral requirements on micro-loans, but instead to

concentrate on close supervision of the MFI’s lending systems and repay-
ment history.
Banks in some places have been required to book a loan-loss provision
expense to cover the full value of uncollateralized loans they make, even
before they become delinquent. Again, this is unworkable for micro-credit.
Even if the provision expense is later reversed when the loan is collected,
the accumulated charges for loans that are showing no problems would
produce a massive under-representation of the MFI’s real net worth. And
such a requirement has no empirical justification in the case of micro-
credit. Thus, general provisions (provisions booked when the loan is made,
and so not based on the presence of any repayment delays) should be no
more stringent for micro-credit than for normal portfolios.
The picture changes, however, once a micro-loan has actually fallen late.
When one narrows the focus down to the micro-borrowers who do run into
repayment problems, experience shows that ultimate collection of their
loans is less likely than collection of collateralized loans that fall late by the
same amount of time. As a result, provisioning of already-delinquent loans
needs to be more aggressive for micro-credit than for normal collateralized
loans.
Loan file requirements

Given the nature of microfinance borrowers, their informal businesses,
and their loan sizes, it would be unreasonable to make micro-lenders gen-
erate the same loan documentation that is required for normal bank loans.
This is particularly true in the case of financial statements for the bor-
rower’s business, evidence that the business is formally registered, or col-
lateral documentation. On the other hand, micro-loan files should always
contain at least the loan contract, a record of the borrower’s repayment
history on prior or concurrent obligations, and a simple estimate of the
borrower’s income, expenses, and repayment capacity, at least where
the MFI’s methodology relies on loan officers rather than fellow group
members to determine repayment capacity. However, MFIs that make
repeated short-term loans, for instance every three months, should not be
required to do a fresh analysis of borrower cash flow before every single
loan.
Restrictions on co-signers as borrowers

Regulations sometimes prohibit a bank from lending to someone who
has co-signed or otherwise guaranteed a loan from that same bank. Such
rules would need to be waived for MFIs that do group lending with cross-
guarantees among the group.
96 Microfinance
Insider lending
Loans made to owners, directors, or managers of a bank are not likely to
receive the same objective scrutiny as loans to unrelated parties. In recog-
nition of this fact, most banking authorities now restrict insider lending to
some limited percentage of the bank’s assets or equity. This author’s view
is that insider lending should be completely prohibited in licensed MFIs,
with the exception of small welfare loan programs for employees. When
specialized MFIs are receiving favorable regulatory treatment for the sole
reason that they are extending financial access to low-income customers, it
is hard to see any reason or need for insider lending.
Frequency and content of reporting

Banks may be required to report their financial position frequently, even
daily. In many settings, the undeveloped state of transportation and com-
munication infrastructure may make this difficult or impossible for rural
banks or branches. In addition, frequent or voluminous reporting to the
banking supervisor can add substantially to the administrative costs of an
intermediary, especially one that specializes in very small transactions. The
chief financial officer of BancoSol once estimated that compliance with the
banking supervisor’s reporting requirements cost the bank 5 percent of its
assets the first year, and 1 percent or more a year thereafter. On the other,
effective supervision is impossible without adequate reporting. Specialized
microfinance banks or branches usually present a less complex set of risks
than normal banking, so it should be possible to supervise them well based
on reporting that is somewhat less burdensome and expensive than what
conventional banks have to provide.
Physical security and branching requirements

Banks’ hours of business, location of branches, and security requirements
are often strictly regulated in ways that could impede service to a
microfinance clientele. For instance, the convenience of clients who are
running their micro-businesses all day might require operations outside
normal business hours, or cost considerations might require that staff
rotate among branches that are open only one or two days a week. Security
requirements such as guards or expensive vaults, or other normal infra-
structure rules, could make it too costly to open small-volume branches
in poor areas. Branching and physical security requirements merit re-
examination—but not necessarily elimination—in the microfinance con-
text. Clients’ needs for financial services have to be balanced against the
security risks inherent in managing cash.
Ownership requirements
Some countries have ownership-diversification rules that prohibit any
single party from controlling more than 20 percent (for instance) of a
bank’s shares. Also, NGOs may not be eligible to own bank shares. Both of
these rules serve legitimate prudential objectives, but they can cause serious
difficulty in the common case where the assets of a newly licensed MFI
come almost exclusively from an NGO that has built up the business over
a number of years. In recent years, commercial and quasi-commercial
investors are showing greater interest in buying shares of newly licensed
MFIs, but there are many transforming MFIs for whom attracting such
investment is not a practical option, at least not at the time that they convert
to licensed form. If the original NGO has to find new owners who will pur-
chase 80 or 100 percent of the business, transformation into licensed form
could be delayed a long time. Some banking supervisors are allowing the
original NGO to own most or all of the shares of a newly licensed MFI,
with a requirement that the ownership structure has to be brought into line
with normal banking rules over a reasonable period of years.
Deposit insurance
In order to protect smaller depositors and reduce the likelihood of runs on
banks, many countries provide explicit insurance of bank deposits up to
some size limit. Some other countries provide de facto reimbursement of
bank depositors’ losses even in the absence of an explicit legal commitment
to do so. There is considerable debate about whether public deposit insu-
rance is effective in improving bank stability, whether it encourages inap-
propriate risk-taking on the part of bank managers, and whether such
insurance would be better provided through private markets. In any event,
if deposits in commercial banks are insured, the presumption probably
ought to be that deposits in other institutions prudentially licensed by the
financial authorities should also be insured, absent strong reasons to the
contrary.
Branchless banking
In a growing number of developing and transition countries, financial ser-
vices are being provided outside of conventional bank branches. The use of
automated teller machines (ATMs) has been spreading for years. More
recently, payment, transfer, and savings services are being offered through
post offices or retail outlets like groceries, pharmacies, or gas stations. Such
services may be used mainly by the middle class, but they hold promise for
poor people as well, especially the rural poor. Using such “retail agents,”
banks can reach places where building and staffing a branch would
be unprofitable because of remoteness, low client density, or low client
98 Microfinance
transaction sizes. In addition, mobile phone operators in countries like the

Philippines, South Africa, and Kenya are exploiting their networks to
provide fast and convenient payment and transfer services to their sub-
scribers, who include increasing numbers of low-income people.
Some branchless banking is bank-led: all of the clients’ transactions are
with a licensed commercial bank, and the retail agents or mobile phone
operators are acting as third-party agents to handle cash on the bank’s
behalf. The bank remains responsible for any cash received. These arrange-
ments raise some regulatory risk, including security of cash-handling and
proper training of agents, but in general they do not add materially to the
risks that are present in normal branch-based banking.
More of a regulatory challenge is presented by non-bank-led models,
where the client’s cash is taken and held by a company like a mobile phone
operator that is not licensed and prudentially supervised by the banking
authorities. When such companies are holding significant amounts of cus-
tomers’ cash, should they be required to meet the same prudential stan-
dards as banks? South Africa’s answer to that question is a conservative
one: any mobile phone operator that wants to provide “e-money” services
is required to partner with (that is, operate under the license of) a commer-
cial bank. This raises the mobile operator’s costs considerably, and these
costs must eventually be passed along to the client.
The Philippines is starting with a more liberal approach, allowing mobile
operators to operate independent of a banking license. However, the
amount of such transactions and the size of outstanding balances owed an
individual customer are capped at low levels. So far this arrangement has
not created significant problems, but the central bank is moving to tighten
restrictions further.
It is not yet possible to identify best practices in dealing with this issue:
so far, the European Community has not been able to agree on a common
approach.
SUPERVISORY ISSUES
The Burden of Supervising Small Intermediaries
For bank supervisors in many developing countries (though certainly not

all), the central fact of life is responsibility for supervising a commercial
banking system with severe structural problems, often including some
sizable banks teetering dangerously close to the edge of safety. The collapse
of one—or a half-dozen—of these banks could threaten the country’s
financial system with implosion. In trying to manage bank risk, supervisors
may have to work in a political minefield, because the owners of banks are
seldom under-represented in the political process. The supervisors’ legal
authority to enforce compliance or manage orderly clean-ups is often inad-
equate. They may not have enough control over the tenure, qualification,
and pay of their staff. Monitoring healthy banks is challenging enough, but
the real problems come when it is time to deal with institutions in trouble.
When a sick bank finally crumbles, its president can start sleeping again
(though perhaps in a different country), while it is the supervisor who has
to stay awake at night worrying.
If bank supervisors sometimes display resistance to adding MFIs—
mostly small, mostly new, mostly weak on profitability—to their basket of
responsibilities, we should recognize that their reasons may be nobler than
narrow-mindedness or lack of concern for the poor.
The Philippines licenses hundreds of small intermediaries as “rural
banks.” Originally, the minimum capital requirement for a rural bank was
very low. These banks are not microfinance institutions, but their opera-
tions do include credit and deposit services for lower-income clients. They
have access to the national payments system and are supervised by the
central bank. As of September 1997, 824 rural banks were serving half a
million clients. These banks had only about 2 percent of the banking
system’s assets and deposits, but they made up 83 percent of the institutions
the central bank had to supervise.
Supervising the rural banks severely stretched the resources of the
central bank’s supervision department, tying up as much as one-half of its
total staff and budgetary resources at times. In the early 1990s one in every
five rural banks had to be shut down, and many others had to be merged
or otherwise restructured. An unpublished 1996 analysis reported that
about 200 inspectors were assigned to the rural banks, but even this level of
resources was viewed as inadequate. Each on-site examination consumed
up to three person-weeks or more. At one point the supervisory department
found that this burden, combined with its budget limitations, was severely
endangering its ability to function.
One of the responses to the crisis was to raise the minimum capital sub-
stantially. But a knowledgeable observer has guesstimated that as of late
2007, perhaps half of the rural banks are still technically insolvent, though
these tend to be the smaller ones. The larger rural banks have most of the
assets and customers are said to be strong and expanding aggressively.
The occurrence of a supervisory meltdown doesn’t necessarily mean that
licensing rural banks has been a failure in the Philippines. Hundreds of
thousands of people are still getting services that would otherwise have
been unavailable to them. But the experience, and similar ones in Indonesia
and Ghana, teaches us to be realistic about the difficulty of supervising
100 Microfinance
large numbers of small new financial institutions. Some would argue that
ineffective supervision is worse than no supervision at all, because it mis-
leads depositors and tarnishes the credibility of the banking authorities.
Minimum Capital as a Rationing Device
When regulators set minimum capital requirements for licensing MFIs, a

major consideration should be limiting new licenses to a number that is
consistent with available supervisory resources. Obviously, this has to be
balanced against the objective of opening access to financial markets, an
objective that tends in the direction of keeping minimum capital as low as
possible. There is a strong argument to be made that regulators should start
with relatively high minimum capital for a new licensing window, and grad-
ually relax the requirement after there has been more supervisory experi-
ence, and supervisors are better able to judge what they can take on.
A country does not necessarily need large numbers of licensed MFIs in
order to serve its market well. In most countries, a few large MFIs account
for the vast majority of the outreach. In 2000, Bangladesh probably had over
1000 MFIs, but the largest ten served all but about 15 percent of the clients.
Small Community-based Intermediaries
Smaller institutions may not require as much supervision as big ones, but
there are lower limits to how far supervision can be watered down. At some
point, “supervision lite” is no longer effective, if effectiveness means that
the supervisor can expect to flag most problems before they have gotten too
serious to fix.
Some member-owned intermediaries take deposits but are so small, and
sometimes so geographically remote, that they cannot be supervised on any
cost-effective basis. This poses a practical problem for the regulator. Should
these institutions be allowed to operate without prudential supervision, or
should minimum capital or other requirements be enforced against them so
that they have to cease taking deposits? Sometimes regulators are inclined
to the latter course. They argue that institutions that cannot be supervised
are not safe, and therefore should not be allowed to take small depositors’
savings. After all, are not small and poor customers just as entitled to safety
as large and better-off customers?
But this analysis is too simple if it does not consider the actual alterna-
tives available to the depositor. Poor people can and do save. If formal
deposit accounts are not available, they have to fall back on savings tools
like currency under the mattress, livestock, building materials, or informal
arrangements like rotating savings and credit clubs. All of these vehicles are
risky, and in many if not most cases, they are more risky than a formal
account in a small unsupervised intermediary. Closing down the local
savings and loan cooperative may in fact raise, not lower, the risk faced by
local savers by forcing them back to less satisfactory forms of savings.
Because of these considerations, most regulators facing the issue have
chosen to exempt community-based intermediaries below a certain size
from requirements for prudential regulation and supervision. The size
limits are determined by number of members, amount of assets, or both.
(Sometimes the exemption is available only to “closed bond” institutions
whose services are available only to members of a pre-existing group such
as employees of a company.) Once the size limits are exceeded, the institu-
tion must comply with prudential regulation and be supervised. If small
intermediaries are allowed to take deposits without prudential supervision,
a good argument can be made that their customers should be clearly
advised that no government agency is monitoring the health of the institu-
tion, and thus that they need to form their own conclusions based mainly
on their knowledge of the individuals running the institution.
Supervisory Tools and their Limitations
Some standard tools for examining banks’ loan portfolios are ineffective for
micro-credit. As noted earlier, loan-file documentation is a weak indicator
of micro-credit risk. In a commercial bank, one can often capture most of
the portfolio risk by examining a small number of large loans, but this is
not true in a micro-credit portfolio consisting of thousands of tiny loans.
Sending out confirmation letters to verify account balances is usually
impractical for micro-credit, especially where client literacy is low. Instead,
the examiner must rely more on analysis of the institution’s lending systems
and their historical performance. Analysis of these systems requires knowl-
edge of microfinance methods and operations, and drawing practical con-
clusions from such analysis calls for experienced interpretation and
judgment. Supervisory staff are unlikely to monitor MFIs effectively unless
they are trained and to some extent specialized.
When an MFI gets in trouble and the supervisor issues a capital call,
many MFI owners are not well-positioned to respond to it. NGOs who own
shares may not have enough liquid capital available. Development agencies
and development-oriented investors usually have plenty of money, but their
internal procedures for disbursing it sometimes take so long that a timely
response to a capital call is impractical. Thus, when a problem surfaces in
a supervised MFI the supervisor may not be able to get it solved by a timely
injection of new capital, as the Colombian banking supervisor found out
when the MFI FinanSol ran into trouble.
102 Microfinance
Another common tool that supervisors use to deal with a bank in trouble
is the stop-lending order, which prevents the bank from taking on further
credit risk until its problems have been sorted out. A commercial bank’s
loans are usually collateralized, and most of the bank’s customers do not
necessarily expect an automatic follow-on loan when they pay off their exist-
ing loan. Therefore, a commercial bank may be able to stop new lending for
a period without destroying its ability to collect its existing loans. The same
is not true of most MFIs. Immediate follow-on loans are the norm for most
micro-credit. If an MFI stops issuing repeat loans for very long, customers
lose their primary incentive to repay, which is their confidence that they will
have timely access to future loans when they want them. When an MFI stops
new lending, many of its existing borrowers will usually stop repaying. This
makes the stop-lending order a weapon too powerful to use, at least if there
is any hope of salvaging the MFI’s portfolio.
A typical MFI’s close relationship with its clients may mean that loan
assets have little value in the hands of a different management team.
Therefore, a supervisor’s option of encouraging the transfer of loan assets
to a stronger institution may not be as effective as in the case of collateral-
ized commercial bank loans.
The fact that some key supervisory tools do not work very well for
microfinance certainly does not mean that MFIs cannot be supervised.
However, regulators should weigh this fact when they decide how many new
licenses to issue, and how conservative to be in setting capital standards or
required levels of past performance for transforming MFIs.
Where to Locate Microfinance Supervision
Given the problem of budgeting scarce supervisory resources, alternatives

to the conventional supervisory mechanisms used for commercial banks
are frequently proposed for depository MFIs.
Within the existing supervisory authority?

The default option for MFI supervision would normally be the supervisory
authority responsible for commercial banks. Using this agency to supervise
microfinance takes advantage of existing skills and lowers the incentive for
regulatory arbitrage. If this option is chosen, the next question is whether
to create a separate department of that agency. The answer will vary from
country to country, but at a minimum, a specially trained supervisory staff
is needed, given the differing risk characteristics and supervisory tech-
niques in the case of MFIs and microfinance portfolios.
The location of microfinance supervision becomes a more compli-
cated question when both non-depository micro-lending institutions and
depository MFIs are to be addressed within a single, comprehensive regula-

tory scheme. The tasks involved in issuing permits to non-depository micro-
lending institutions have relatively little to do with the prudential regulation
and supervision of licensed depository institutions. In some contexts,
lodging both of these disparate functions within the same regulatory body
might be justified on pragmatic grounds—such as the absence of any other
appropriate body, or the likelihood that the permit-issuing function would
be more susceptible to political manipulation and abuse if carried out by
another body. In other cases, non-depository MFIs are required to report to
the banking supervisor in order to make it easier for them to move eventu-
ally into more services and more demanding prudential regulation. Often,
however, the risks of consolidating prudential and non-prudential regulation
of microfinance within the banking authority will outweigh the benefits.
These risks include the possibility of confusion on the part of supervisors as
to the appropriate treatment of non-depository institutions, and the possi-
bility that the public will see the supervisory authority as vouching for the
financial health of the non-depository institutions, even though it is not (and
should not be) monitoring the health of these institutions closely.
A separate and independent agency?

In some countries the banking authorities’ reluctance to take responsibility
for microfinance leads to plans or decisions to lodge microfinance super-
vision in an independent agency. Building skills and experience in a different
body can be time-consuming, and the new MFI supervisor may not be as
politically independent as the banking supervisor. One approach to short-
ening the learning curve is to entrust prudential supervision of deposit-
taking MFIs to an apex agency that is already making wholesale loans to
MFIs. This structure can present conflicts of interest. If such an apex super-
visor has large loans outstanding to a troubled MFI, will the agency be
tempted to drag its feet when depositors’ interests are best served by shut-
ting the MFI down? On the other hand, central banks frequently deal with
similar conflicts of interest.
In developing countries, regulation and supervision of credit unions has
usually been lodged outside the banking authority, often in the government
department that is responsible for cooperatives of all sorts. The experience
with this arrangement has usually been very disappointing.
Delegated supervision?
Sometimes the government financial supervisor delegates part or all of the
tasks of direct supervision to an outside body, while monitoring and con-
trolling that body’s work. This seems to have worked, for a time at least, in
some cases where the government financial supervisor closely monitored
104 Microfinance
the quality of the delegated supervisor’s oversight, although it is not clear

that this model reduces total supervision costs. Where this model is being
considered, it is important to have clear answers to three questions: (1) who
will pay the substantial costs of the delegated supervision and the govern-
ment supervisor’s oversight of it? (2) if the delegated supervisor proves
unreliable and its delegated authority must be withdrawn, is there a realis-
tic fallback option available to the government supervisor? and (3) when
a supervised institution fails, which body will have the authority and
resources to clean up the situation by intervention, liquidation, or merger?
Because many MFIs are relatively small, there is some temptation to
think that their supervision, or at least on-site inspection, can be safely del-
egated to external audit firms. Unfortunately, experience has been that
external audits of MFIs, even by internationally affiliated audit firms, very
seldom include testing that is adequate to provide a reasonable assurance
as to the soundness of the MFI’s loan assets, which is by far the largest risk
area for micro-lenders. If reliance is to be placed on auditors, the supervisor
must require microfinance-specific audit protocols that are more effective,
and more expensive, than the ones now in general use, and must regularly
test the auditors’ work.
Self-regulation and supervision

When regulators decide that it is not cost-effective for the banking author-
ity to provide direct oversight of large numbers of MFIs, self-regulation is
sometimes suggested as an alternative. Discussion of self-regulation tends
to be confused because people use the term to mean different things. In this
discussion, “self-regulation” refers to regulation (and/or supervision) by
some body that is effectively controlled by the regulated entities.
This is one point on which historical evidence seems clear. Self-regulation
of non-bank financial intermediaries in developing countries has been tried
many times, and has virtually never been effective in protecting the sound-
ness of the regulated organizations. One cannot assert that effective self-
regulation in these settings is impossible in principle, but it can be asserted
that such self-regulation is almost always an unwise gamble against very
long odds, at least if it is expected that the regulation and supervision actu-
ally enforce financial discipline and prudent risk management. Sometimes
regulators have required certain small intermediaries to be self-regulated,
not because they expect the regulation and supervision to be effective, but
because this is politically more palatable than saying that these deposit-
takers will be unsupervised. This can be a sensible accommodation in some
settings. While self-regulation probably will not keep financial intermedi-
aries healthy, it may have some benefits in getting institutions to begin a
reporting process, or in articulating basic standards of good practice.
AFTERWORD: DOES PRUDENTIAL REGULATION

WORK?
There has long been a respectable minority of academic opinion that is skep-
tical about the conventional wisdom that government prudential regulation
and supervision are effective in limiting bank failures and financial system
crises. Barth, Caprio and Levine (2006) conducted a cross-country analysis
based on a 150-country database, and drew some jarring conclusions:
In terms of what works best, our analyses raise a cautionary flag regarding the
foundations of current international best practice recommendations. In par-
ticular, our results question the efficacy of Basel II’s first two pillars on capital
regulations and official supervision . . .
Across the different statistical approaches, we find that empowering direct

official supervision of banks and strengthening capital standards do not boost
bank development, improve bank efficiency, reduce corruption in lending, or
lower banking system fragility. Indeed, the evidence suggests that fortifying
official supervisory oversight and disciplinary powers actually impedes the
efficient operation of banks, increases corruption in lending, and therefore hurts
the effectiveness of capital allocation without any corresponding improvement
in bank stability . . .
In contrast to these findings on capital regulations and direct supervisory over-

sight of banks . . . supervisory and regulatory policies that facilitate private
sector monitoring of banks improve bank operations, which endorses Basel II’s
third pillar on market discipline. One mechanism for fostering private monitor-
ing of banks is by requiring the disclosure of reliable, comprehensive, and timely
information. Countries that enact and implement these pro-private monitoring
regulations enjoy more efficient banks and suffer from less corruption in lending.
Furthermore, laws that strengthen the rights of private investors enhance
the corporate governance of banks. In contract, policies (for example, deposit
insurance) that weaken market monitoring of banks tend to have adverse
ramifications on the banking system [including reducing system stability rather
than strengthening it] . . .
We recognize, of course, that many countries do not have the legal and political
institutions necessary to support effective market monitoring of banks. Con-
sequently, many readers may conclude that a practical approach involves
empowering official supervisors until countries develop the institutional foun-
dations for market monitoring. The cross-country results thus far, however, do
not support this conclusion. The results instead indicate that regulatory restric-
tions on bank activities, impediments to the entry of new banks, government
ownership of banks, and reliance on powerful official supervisors to oversee
banks have adverse effects on the operation of banks. Moreover, it is exactly in
countries with weak political and legal institutions that empowering official
supervisors is likely to be most detrimental. (Barth, Caprio and Levine, 2006,
pp. 10–16)
106 Microfinance
Other econometricians disagree about interpretation of the data, and

this author has neither the competence nor the courage to offer any judg-
ment about the issue. In any event, hardly any of the world’s countries have
been willing to take the advice of those who favor eliminating official super-
vision of banks. Whatever the merits of the issue, it would be politically
impossible for most governments to back away from protecting depositors.
Then why raise the question in a discussion about regulating micro-
finance? The reason is that in crafting a regulatory regime for microfinance,
many decisions about detail and degree depend on one’s underlying
assumptions about the appropriateness and effectiveness of government
prudential oversight. To the extent that one has doubts about such over-
sight, or is concerned that regulatory power put into the hands of a gov-
ernment supervisor may be used for personal rather than public benefit,
then one might be more cautious in deciding particular issues related to the
scope of such power.
NOTES
1. Author’s note: this chapter draws heavily on a set of consensus guidelines developed in
consultation with a wide range of experienced regulators, policy advisors, and industry
analysts (Christen, Lyman and Rosenberg, 2003) as well as the working experiences of the
author and many colleagues. Citation of sources will be limited. The discussion is meant
to apply to developing and transition economies only: the dynamics of microfinance in
rich countries, and the regulatory issues they present, can be substantially different.
2. It is important to note that if Compartamos priced its loans to produce no profit at all, it
would still have to charge interest of about 77 percent, which would shock most con-
sciences despite being completely driven by costs of lending. Compartamos’s loans are
extremely small in relation to the Mexican context, so its administrative costs alone are
very high (Rosenberg, 2007).
3. In most countries, “micro-credit” is understood to refer mainly to small uncollateralized
loans to people who have informal micro-businesses rather than jobs in the formal sector.
4. This section draws heavily on Isern, Porteus, Hernandez-Coss and Egwuagu (2005).
5. www.mixmarket.org/en/demand/demand.show.profile.asp?ett=1784&.
REFERENCES
Barth, J.R., G. Caprio and R. Levine (2006), Rethinking Bank Regulation: Till
Angels Govern, New York: Cambridge University Press.
Christen, R.P., T.R. Lyman and R. Rosenberg (2003), “Guiding principles on
regulation and supervision of microfinance”, CGAP, accessed at www.cgap.
org/portal/binary/com.epicentric.contentmanagement.servlet.ContentDelivery
Servlet/Documents/Guideline_RegSup.pdf.
Isern, J., D. Porteus, R. Hernandez-Coss, and C. Egwuagu (2005), “AML/CFT
regulation: implications for financial service providers that serve low-income
people”, CGAP Focus Note No. 29, accessed at. www.cgap.org/portal/binary/
com.epicentric.contentmanagement.servlet.ContentDeliveryServlet/Publication
s/html_pubs/FocusNote_29.html.
Rosenberg, R. (2007), “CGAP reflections on the Compartamos initial public
offering: a case study on microfinance interest rates and profits”, CGAP,
Focus Note No. 42, accessed at www.cgap.org/portal/binary/com.epicentric.con-
tentmanagement.servlet.ContentDeliveryServlet/Documents/FocusNote_42.pdf.
6. Gender empowerment in
microfinance
Beatriz Armendáriz1 and Nigel Roome
INTRODUCTION
Ever since microfinance was popularized in the mid-1970s in Bangladesh,

one of its salient features has been the overwhelming representation of
women. The trend has increased steadily, particularly during the 1980s.
According to 2006 Microcredit Summit Campaign report, seven out of
ten microfinance clients are women.2 Millions of these women are married
or live with a partner, and many have children. Relative to initial lending
practices by the Grameen Bank in Bangladesh, the bias in favor of loans
to women in microfinance has been accompanied by an increasing trend
to exclude men from microfinance services, particularly at very low
income levels. The practice of exclusion might however prove to be coun-
terproductive, for it can generate frictions within households, as men feel
increasingly threatened in their role as primary breadwinners within the
household.3
In this chapter we argue that the promotion of women in microfinance
initiatives and the bias against men is taking place in the absence of solid
empirical evidence on the effects of this strategy, and on the balance of
power in households and on the health, education, and well-being of all
household members, which we hold to be key aspects of development. We
further argue that this issue deserves research given the possibility of
unforeseen outcomes and adverse consequences that run counter to the
goal of microfinance initiatives to promote development.
To clarify the central issues, on the one hand, higher household income
in the hands of women might increase health and education for women and
their household members—we call this the women-empowerment effect.
On the other hand, the exclusion of men from access to subsidized finance
might create frictions, and rebound effects, that diminish the supportive
role women play for their spouses and wider household members in the pro-
duction of health and education—we call this the women-disempowering
effect. In the event that the latter effect dominates over the former, then
108
Gender empowerment in microfinance 109
subsidized microfinance for women might have no impact, or even worse, a

negative impact on health and education. An even more challenging ques-
tion is what social and institutional conditions most strongly influence
these empowerment and disempowerment effects and their outcomes.
This chapter is structured as follows. First, it provides an overview of
what we currently know about microfinance, gender, health, and education
in the context of Bangladesh, where most research has been conducted.
Second, some anecdotal evidence from Bangladesh and Africa on the
notion of microfinance empowerment is presented and discussed. This
raises questions about the structures on the enhanced capacity of women
to assert their role as the main providers of health and education, mainly
arising from the fact that the empowerment of women generates frictions
with their partners, which in turn leads to a potential disempowerment
effects. Third, anecdotal evidence from Chiapas in southern Mexico is out-
lined, which provided the basis for empirical research on new approaches
to microfinance now being undertaken in the region. Fourth, the chapter
outlines this experimental intervention in southern Mexico, where the
women borrowers in a microfinance initiative invited their spouses to be
part of women-only solidarity groups as borrowers, in order to see whether
potential frictions could be eliminated as a way to enhance women empow-
erment and provide for better access to health and education at the house-
hold level. The main challenges of implementing this type of intervention,
which were revealed through this empirical study so far are described. The
final section spells out some concluding remarks.
CURRENT KNOWLEDGE OF MICROFINANCE,

FINANCIAL RESOURCES, AND GENDER AS A BASIS
FOR THE PRO-WOMEN BIAS4
The most influential empirical study on microfinance and gender can be

found in an article published by the Journal of Political Economy in 1998
by Mark Pitt and Shahidur Khandker. In their study, Pitt and Khandker
develop a framework for estimating the impact of microfinance using
cross-section data from Bangladesh for 1991–92. The paper pins down the
potential sources of bias, in identifying and estimating the impact of
microfinance initiatives alone on outcomes such as household expenditures
on health and education.
For example, Pitt and Khandker addressed the bias that might arise
because the individuals who self-select into microfinance programs may be
the least poor and most entrepreneurial members of their community. This
bias would lead to an overestimate of the overall potential of microfinance
110 Microfinance
on poverty reduction as and when microfinance projects were expanded to

include poorer and less entrepreneurial individuals. Pitt and Khandker
faced the well-known endogeneity problem that entrepreneurial individ-
uals may deliver for themselves better incomes, which then enable them to
qualify for further loans, which in turn increase their incomes. Typical ways
to resolve this problem of estimating is through the use of an independent
variable, which correlates with entrepreneurial activity but not with out-
comes. Pitt and Khandker used land-ownership as an instrumental vari-
able: to qualify for a microfinance loan, individuals (both men and women)
had to be poor as proxied by their holdings of land not being more than
half an acre. They used in this instrumental variable in studies that
compared villages with microfinance opportunities and control villages
without. This approach meant that those who received loans in treatment
villages did so because they were landless poor, with the same entrepre-
neurial abilities as those in the control villages where the landless poor
did not access microfinance loans because there were not that many
microfinance providers. Once village characteristics were controlled for, Pitt
and Khandker extended their analysis to the role of gender.
In particular, taking advantage of the fact that Bangladeshi microfinance
enterprises at the time were lending to men as well as women, Pitt and
Khandker conducted a study to assess the relative impact of lending to men
head-of-households compared with their women counterparts on develop-
ment outcomes in terms of health and education. Their results regarding rel-
ative provision for health and education by both types of household heads
were, at best, unclear. However, the main findings of the study of expendi-
tures by both types of households in relation to household expenditures (par-
ticularly on food and tools for expanding their respective businesses) are well
known. These results have been highly influential, particularly in shaping the
trend in focusing donors’ aid through subsidized loans toward women.5
In particular, Pitt and Khandker showed that when a loan of 100 taka
was extended to men it translated into 11 taka going into household expen-
ditures (for food/nutrition/working tools), while the same amount lent to
women household heads led to 18 taka being spent on household expendi-
tures (for food/nutrition/working tools).
While it would be too bold to claim that the findings of Pitt and
Khandker alone have influenced the bias to women in recent microfinance
initiatives. It is our conjecture that in the absence of any countervailing
empirical evidence, Pitt and Khandker’s findings contributed to the norms
and operational practices of CGAP (Consultative Group to Assist the
Poor) World Bank, as well as many other multilateral organizations
engaged in providing subsidized microfinance. Their priority has been to
direct subsidized loans to women.
The common practice in favor of subsidized micro-loans to women also

flows from the research on the practice of delivering aid to women. For
example, food stamps in the United Kingdom and Sri Lanka, and staple
food and cash deliveries under the PROGRESA (now called OPORTU-
NIDADES) program in Mexico were directed to women rather than their
spouses. This was done for fear that if such aid was given to men, they
might sell the food stamps and mis-spend the resources, possibly wasting
money on gambling, tobacco, and alcohol.
There are a number of empirical studies on the practice of targeting aid
to women. Emmanuel Skoufias (2001) reports that the OPORTU-
NIDADES project aimed at women in rural Mexico led to sharp social
improvements: poverty decreased by 10 percent, school enrollment
increased by 4 percent, food expenditures increased by 11 percent, and
adults’ health (as measured by the number of unproductive days due to
illness) also improved considerably.6
Duncan Thomas (1990) reports that child health in Brazil (as measured
by survival probabilities, height-for-age, and weight-for-height) along with
household nutrient intakes, tended to rise if additional non-labor income
was in the hands of women rather than men. He observed that income in the
hands of a mother had, on average, 20 times the impact of the same income
in the hands of a father with respect to children’s survival probabilities. In a
subsequent study, also on Brazil, Thomas (1994) reports that increasing the
bargaining power of women is associated with increases in the share of the
household budget spent on health, education, and housing as well as
improvements in child health. Patrice Engle (1993) similarly studies the rela-
tionship between a mother’s and father’s income on child nutritional status
(height-for-age, weight-for-age, and weight-for-height) for hundreds of
households in Guatemala, and reports that children’s welfare improves as
women’s earning power increases relative to their husbands’. Paul Schultz
(1990) finds that in Thailand non-labor income in the hands of women tends
to reduce fertility more than non-labor income possessed by men. He also
finds that the impact of non-labor income has different effects on labor
supply, depending on which household member controls that income.7
Anderson and Baland’s (2002)’s article on Rotating Savings and Credit
Associations (ROSCAs) reports on a survey of hundreds of women in
Kenya. An overwhelming majority of the women responded that the prin-
cipal objective for joining a ROSCA was to save money, and nearly all of
the respondents were married. Anderson and Baland conclude that an
important motive for women joining ROSCAs is thus the desire to keep
money away from their husbands. Other studies, not necessarily confined
to ROSCAs, suggest that savings motives (and the protection of assets) also
apply to women’s involvement in microfinance institutions.
112 Microfinance
Christopher Udry’s (1996) research on agricultural practices in Burkina

Faso provides evidence on the ways men and women invest in agriculture.
Using panel data, controlled for soil quality and other variables, he finds
that agricultural productivity is higher in plots cultivated by men than
women. He also finds that compared with plots cultivated by women, the
higher yields of plots cultivated by men are due to a greater intensity of
productive inputs (including fertilizer and child labor). He thus concludes
that productivity differentials are attributed to the intensity of production
between plots cultivated by men and women and not to inherent skill
differentials: an outcome he regards as inefficient since there are sharply
diminishing returns to the use of fertilizer. Not only are resources not
fully shared, they are allocated in ways that diminish total household
income. Udry suggests that by reallocating inputs to plots cultivated
by women can thus enhance efficiency. Another solution (that is, the
microfinance solution) is to provide women with credit sufficient to pur-
chase additional inputs. A second way that microfinance can potentially
address problems like this is by tackling the social norms that prevent
women from having adequate access to inputs and marketing facilities in
the first place. This might be done through demonstration effects and
from pressure created by the micro-lender to ensure higher returns to bor-
rowers’ investments.
From the point of view of practice, a field loan officer would see women
as better customers for loans compared with men for at least four reasons.
First, repayment rates on loans by women are higher, because women are
more risk-averse and therefore more conservative in their investment strat-
egy. Also, women are more vulnerable to peer pressure and the threat of
public humiliation with regards to failure in the repayments on their loans,
women have less opportunities than men to access alternative sources of
credit, which in turn reduces the scope for moral hazard.8 Moreover, field
practitioners in microfinance argue that women are less argumentative,
which reduces the transaction costs of the loan, both for their peers and the
bank. Women also lower the agency costs of bank officers because as
groups they are more punctual at repayment meetings, which avoids the
bank officer having to devote time looking for them at their homes/
businesses. Last but not least, women loan officers cost less than men, and
in many instances women are more efficient at granting and collecting
repayments.9
Taken together, the findings of empirical investigations, the perspectives
of donors, and experience of practitioners, have led to an established
wisdom in favor of lending to women. Moreover, the conventional wisdom
has been that excluding men from microfinance has no significant or im-
portant detrimental outcomes. However, more recent views from the field
expressed at the recent Microfinance Forum in Beijing (2006) suggest

otherwise:
male exclusion can lead to negative consequences for women who join financial
services: they may meet resistance from men who see their exclusive participa-
tion as unfair and threatening; their loans may be hijacked . . . A family whose
adult members all have access to financial services is better off than one where
half are ineligible. (Hugh Allen at the Microfinance Forum, 2006)
While the experiential knowledge of people like Hugh Allen should not be
accepted without detailed investigation, his views have been a consideration
for social scientists and anthropologists voicing similar concerns for some
time now. Their observation, which run counter to conventional wisdom
are reviewed in the following section.
ANECDOTAL EVIDENCE FROM BANGLADESH AND

AFRICA
In this section, we argue that there are potential dangers in excluding men
from subsidized microfinance as this may lead to frictions between house-
hold heads, leading to lower quality and quantity of health and education
provision within the overall household. At this stage the evidence for this
position is anecdotal, deriving from Bangladesh and Africa. It suggests that
there is a need to take into account the potential danger of excluding the
male head of household from microfinance, as their exclusion can over-
burden women and lower health and education outcomes.
Long before the 2006 Nobel Peace Prize was awarded to the creator of
Grameen Bank, Muhammad Yunus, for his work in microfinance, house-
hold surveys from Bangladesh, dating back to 1999, documented evidence
that microfinance was increasing frictions between husbands and wives, as
husbands often felt threatened in their role as primarily income earners
(Rahman, 1999). Moreover, well-known evidence, also from Bangladesh,
suggests that microfinance does not increase women’s bargaining power
entirely, because on average, women borrowers surrender nearly 40 percent
of their control over the investment decisions they make. More alarmingly,
over 90 percent of the returns these women realize from their investments
are handled by their husbands (Goetz and Sen Gupta, 1996).
In Africa, Linda Mayoux (1999), reports on a survey of 15 different
microfinance programs. She finds that the degree of women’s empowerment
is household- and region-specific, with women’s empowerment dependent
on inflexible social norms and traditions. These findings have to be weighed
against the fact that impact on empowerment will also depend on how well
114 Microfinance
particular programs were designed. The issue of context factors and

program design leads to the exceedingly preliminary observations from a
field experiment undertaken in southern Mexico.
ANECDOTAL EVIDENCE FROM FIELD

EXPERIMENTS IN SOUTHERN MEXICO10
Grameen Trust Chiapas, AC (henceforth called GTC) is one of the first
replications of the Grameen model of microfinance in Latin America. The
project is located in the highlands of southern Mexico. It deploys funds
from the Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) via
Grameen Trust Bangladesh. The replication in southern Mexico started by
lending to women-only groups in 1997.11 In 2003, in sharp contrast to the
original Grameen model, GTC took the risk of lending to men of some
otherwise women-only groups. Since then the organization grew rapidly,
and it now has over 12 000 borrowers in different groups, a large majority
of those mixed groups of women and men.
When branch managers in different geographical locations in the south-
ern Mexican replication are asked why they have accepted men into women-
only groups, four explanations are offered. The first relates to informational
asymmetries between men and women. One loan officer argues that even if
loan disbursements and repayments are publicly known in women-only
groups, men tend to overestimate the amount of money that women are
handling, and they therefore contribute less to overall household expendi-
ture, which often creates frictions within the household. This has dynamic
effects. In many instances, women under these conditions are no longer
using their loans for investment but for normal household expenditures on
food, health, and education (particularly in the month of August when the
academic year starts). They also often quarrel with their husbands who are
no longer providing as much for these expenditures as they used to. Inviting
some men to join the group allows them to have a more accurate estimate
of women’s real investments and their realized returns. With this informa-
tion they are less likely to reduce their contributions to household expen-
ditures. In those groups that became mixed, women borrowers invest more,
and there were increased repayment rates by both men and women in such
groups.
A second explanation relies on the potential workload externalities of
having women as the only recipients of loans within the household. In par-
ticular, another loan officer argues that when women contract a loan from
GTC, they become busier, and that the quality of the services that women
traditionally provide to the household such as meals, and household
chores, decreases in quantity or quality or both. This, the bank manager

argues, irritates men and creates a “tense” atmosphere within the house-
hold. This family tension causes women to default more often or prevents
them from making their repayments on time. When men are invited to join
groups, they seem to internalize the negative workload externalities created
by GTC micro-loans to women. In loan officer Regis Ernesto Figueroa’s
own words: “invited men help more their spouses in their businesses and in
household chores, which in turn, reduces tensions, and enables women to
repay on time, as men become de-facto business partners of women.”
A third explanation relates to the absence of secure places for women to
hide money while they save for two consecutive weeks or more in order to
make their repayments to GTC. In particular, another loan officer argues
that women cannot open bank accounts in commercial banks as these
banks do not accept their very small savings, as the transactions costs for
the commercial banks are too high relative to the amounts deposited.
Women borrowers of GTC therefore hide money away from their husbands
in different places, generally in the house, because husbands steal the
money and use it on alcohol and tobacco. When men are invited to join the
group, this particular loan officer argues, the situation changes because
under the “Grameen rules,” he becomes responsible for the debt of the
other male and women members of the group. “Women become happier.
They no longer complain about their husbands or men in general. The
household heads work harmoniously together,” the loan officer explains.
A loan officer at the headquarters of GTC offers a fourth, and last, expla-
nation of why men are invited into women’s groups. She argues that the
inclusion of men brings more women clients into the scheme, in particular
more single women. She explains that the reason is because women gener-
ally face a trade-off between being financially independent via a micro-loan
from GTC, or, getting married. The argument goes that since GTC accepts
men, women no longer face this trade-off, and they are therefore more likely
to become clients. Moreover, the inclusion of men, according to this loan
officer, has increased marriage rates!
ATTEMPTS TO MEASURE EMPOWERMENT AND

DISEMPOWERMENT12
The anecdotal evidence set out above suggests a substantive need to explore
in greater depth the relationship between microfinance structures and the
issues of gender in development and empowerment around microfinance.
This calls for experiments designed to test the effects of the inclusion
of male heads of households into women-only solidarity groups. Such
116 Microfinance
experiments are exceedingly demanding. Nevertheless they are important

given the challenges to the conventional wisdom, that women are increas-
ingly empowered by microfinance that enables them to expand their busi-
nesses, earn a higher return so that their spouses would value them better,
which translates into higher health and education for the household.
In designing such studies it is recognized that cultural and institutional
differences may impact the results. Ideally, any test of the empowering-
disempowering hypothesis should therefore take place in Bangladesh,
Africa, and Latin America to establish whether the results are culturally
and institutionally robust. However, finding partner microfinance institu-
tions that would allow researchers to conduct scientific experiments of this
kind, is difficult enough; to do so in three continents is yet more difficult.
We report here progress to date with a pioneering study developed by
Harvard and Yale researchers from the Innovations for Poverty Action
(IPA) in their continuing study on the impact of gender issues of
microfinance and health and education outcomes in association with the
Grameen Trust Chiapas, AC in southern Mexico. The elements of this
study are reported below.
Innovations for Poverty Action (IPA) researchers designed a survey and
a follow-up random experiment, using a sample of approximately 2000
borrowers in women-only solidarity groups in 2006. In this experiment
married-women-only solidarity groups were (randomly) selected into treat-
ments and controls. Control solidarity groups were not subject to any kind
of “intervention” while the treatments were.
The treatments were divided into four different sub-groups. Intervention
in the first sub-group consisted of allowing women to voluntarily invite
their partners/husbands to join the Grameen-style solidarity group in order
to acquire a micro-loan. The study sought to take into account the poss-
ible “network effects” that might follow as invited male spouses joined soli-
darity groups, increasing the synergies through the span of group members.
IPA researchers therefore allowed for a sub-group of women who could
invite other women friends to join their group. Similarly, it was recognized
that as partners/spouses were invited to join, so household income was
increased. This factor was taken into account by extending larger loan sizes
to a sub-group consisting of women-only clients. Last but not least, a treat-
ment sub-group consisting of women who could invite their partners/
husbands via providing them high and low monetary incentives to better
proxy women’s marginal benefit from being financially independent.
The IPA researchers are using a follow-up survey of the four sub-
treatment groups and the control group to assess and evaluate any behav-
ioral changes at household level. Some of the questions that will be resolved
are: did women borrowers decide to invite their male spouses, and if so, did
their willingness to do so increase as they were provided with incentives? In

what way did the inclusion of male spouses alter outcomes in terms of
health, education, and child labor, for example? It is recognized that there
are many other possible dimensions of change to address, but it is not pos-
sible to develop a comprehensive questionnaire until the initial results from
the 2008 follow-up survey of behavioral changes has been processed and
analyzed.
In the meantime, IPA researchers have been able to detect some interest-
ing idiosyncrasies in the sample of borrowers, which include five geo-
graphical areas, quite distant from one another. These preliminary findings
show that most of the decisions regarding investment and the expenditures
from returns realized from micro-loans are taken by men heads of house-
holds, not by women, in the bank branch with the highest proportion of
indigenous population. In addition, Grameen Trust Chiapas, as well as
other group-lending institutions such as AlSol, another Grameen replica-
tion founded by Beatriz Armendáriz, have been serving borrowers in these
branches/regions for a long time. Health and educational expenditures in
these two branches do not differ considerably. If microfinance to women
had not already empowered women in such traditional societies, and
assuming the head of household relationship is basically frictionless as
wives systematically defer decision-making to their husbands, it is difficult
to imagine what an intervention of the sort we undertook in that region
could actually bring about in terms of changes in behavioral patterns.
However, when the women in these branches are given the power to invite
their spouses into the group this provides an empowering tool in its own
right, and, household income is expected to increase when women actually
decide to include their spouses. On the other hand we might expect that if
any anticipated increase in household income, once partners are included
as microfinance clients, is then controlled purely by men, there would not
be the expected changes in outcomes, particularly with respect to health
and education. Anticipating this, women might decide not to invite their
spouses in the first place. It is, however, much too early in the experiment
to make any predictions of substantive outcomes from the project.
A somewhat similar scenario seems to prevail in two more affluent
branches and regions. Interestingly, in at least one of the two branches,
women borrowers have remained with the Grameen Trust Chiapas for a
much longer period of time compared with the other four branches. At this
time their higher income and expenditure, on average, might be due to this
continued microfinance activity, nevertheless, educational levels seem to be
just as low as in the poorest branch, while it appears that health expendi-
tures are somewhat higher. Whether women will opt for bringing their
spouses in to the project, and whether this will translate into higher
118 Microfinance
outcomes in terms of health and education remains an open question at

this stage in the program.
An interesting situation exists in the other two branches and regions
where the income of the borrowers is the highest. Women in both these
regions are not just wealthier, but are also more educated and their health
expenditures appear to be higher. Women seem to be more empowered in
that they often declared themselves as being the main household head, and
take most of the household decisions. Their spouses seem to be more sup-
portive of their micro-businesses. In this situation these women might value
their financial independence, and this might be leading to a completely fric-
tionless relationship, in which case we should probably not expect those
already empowered women to actually invite their partners to join their
project either. Again, it is too early to tell as we do not yet know the level
of take-up of male membership, and if the invitation of spouses to join will
lead to higher outcomes in terms of expenditures on health or education.
CONCLUDING REMARKS
At present the baseline survey in Chiapas indicates that the degree of

women empowerment is in line with Linda Mayoux’s (1999) findings in 15
different microfinance programs in Africa. That is to say, expenditure is
household- and region-specific and inflexible as a consequence of social
norms that seem exceedingly difficult to change.
However, empowering women via an additional tool, namely by giving
them the right to voluntarily invite their partners, might help to acceler-
ate the process for change in those social norms. This might, however,
prove more difficult in poorer regions where household heads seem mostly
to be men. The question then is why should subsidized loans that make
women responsible for repayment, but do not give them power over
crucial decisions regarding their business and household expenditures in
health and education seem to be endorsed by donors? Moreover, as the
microfinance industry becomes increasingly commercial, micro-credit
becomes increasingly burdensome on women. Why should women take
on the responsibility over higher repayments in the first place? This view
accords well with for-profit microfinance enterprises in Latin America
where men are increasingly self-selecting into programs offered by such
enterprises. In the absence of subsidies, Grameen Trust Chiapas as well as
other organizations in the region might be increasingly attracting men,
not women. And the interest rates charged could be “friendlier” to
women, if only because women are the main brokers of health and edu-
cation within the household.
As far as the more affluent clients served by Grameen Trust Chiapas

are concerned, it might be that the whole idea of excluding husbands can
be counterproductive, because of informational asymmetries that appear
to lead to mistrust, frictions within the couple, and worse, a decreased
participation by men in household expenditures. This outcome is not
what we understand as a preferred outcome for women. However, such
disempowerment effects should be weighed against the value that women
attach to their financial independence. We see an important balance
between greater financial independence on one side, and more demands
on time and loss of money for established levels of expenditure on the
other.
Given these scenarios women might be reluctant to invite their part-
ners into their groups, but for different reasons. In the case of less affluent
households in very traditional societies, because having partners join the
project would not change anything. And in the case of relatively more
affluent households because women attach too much value to their financial
independence.
Final results on take-up as well as potential behavioral changes from
this experiment on gender should further clarify the questions and shed
light on implications of microfinance structures around the exceedingly
important issues of gender in the field of microfinance, development, and
empowerment.
NOTES
1. I gratefully acknowledge the support of Alissa Fishman, Randall Blair, and Julio Luna
from IPA in Mexico. Co-authors Dean Karlan and Sendhil Mullainathan have been
incredibly patient in teaching me how to conduct field work and without their guidance,
my part in this chapter would never have been written. Valuable comments from seminar
participants at Harvard, Columbia, Solvay Business School, and CERMi in Brussels are
also greatly appreciated. Finally, the collaboration of Grameen Trust Chiapas manage-
ment, and in particular, the support from Ruben Armendáriz, Maricela Gamboa,
branch managers, and loan officers from Tuxtla, Ocosingo, Comitan, and Las
Margaritas, have been exceedingly valuable. Nigel and I are solely responsible for the
views and errors in this chapter.
2. Daley-Harris, Sam (2003).
3. Some evidence on this and follow-up debate is found in Mayoux (1999) and Rahman
(2001), among others.
4. This section borrows from Armendáriz’s joint work with Jonathan Morduch (2005). For
a more general survey on gender issues in economic development, see Duflo (2005).
5. Pitt and Khandker’s econometric estimations and results are, however, exceedingly con-
troversial. For a more comprehensive critique, see Pit (1999) and Armendáriz and
Morduch (2005).
6. Promoting women to powerful positions in villages and regions may, by the same token,
bring social benefits. In a recent paper on India, Raghabendra Chattopadhyay and
Esther Duflo (2003) show that by empowering women, and, in particular, by allowing
120 Microfinance
them to be elected to local councils, spending on public goods most closely linked to
women’s concerns increased.
7. Evidence from India also shows that there is a positive correlation between the relative
size of a mother’s assets (notably jewelry) and children’s school attendance and medical
attention (Duraisamy and Malathy, 1991; Duraisamy, 1992).
8. Hossain (1988): 81 percent of women had no repayment problems versus 74 percent of
men.
9. Khandker, Khalily, and Kahn (1995): 15.3 percent of male borrowers were “struggling”
in 1991 versus 12.4 percent of female (missing some payments before the final due date).
10. This section draws from current field work with Dean Karlan and Sendhil Mullainathan
in southern Mexico.
11. A year later, and under the auspices of Grameen Foundation USA, some of the
Grameen Trust Chiapas’s managers founded AlSol, which currently serves approxi-
mately 3000 borrowers.
12. For an explanation on random experiments, see Duflo, Glennester and Kremer (2006).
REFERENCES
Anderson, Siwan and Jean-Marie Baland (2002), “The economics of ROSCAs and
intrahousehold allocation”, Quarterly Journal of Economics, 117 (3), 983–95.
Armendáriz, Beatriz and Jonathan Morduch (2005), The Economics of
Microfinance, Cambridge, MA: MIT Press.
Chattopadhyay, Raghabendra, and Esther Duflo (2003), “Women as policy makers:
evidence from an India-wide randomized experiment”, typescript, Cambridge,
MA: MIT Economics Department.
Daley-Harris, Sam (2003), State of Microcredit Campaign Report 2003, November,
Washington, D.C.: Microcredit Summit, accessed at www.microcreditsummit.
org/pubs/reports/socr/2003/socr 03_en.pdf.
Duflo, Esther (2005), “Gender equality in development”, Massachusetts Institute
of Technology Poverty Action Lab working paper, Cambridge, MA.
Duflo, Esther, Rachel Glennester, and Michael Kremer (2006), “Using randomiza-
tion in development economics research: a toolkit”, Massachusetts Institute of
Technology Poverty Action Lab working paper, Cambridge, MA.
Duraisamy, Paul (1992), “Gender, intrafamily allocation of resources, and child
schooling in South India”, Yale University Economic Growth Center, working
paper no. 667, New Haven, CT.
Durisamy, Malathy (1998), “Children’s schooling in rural Tamil Nadu: gender dis-
parity and the role of access, parental and household factors,” Journal of
Educational Planning and Administration, XII (2), 131–54.
Engle, Patrice (1993), “Influences of mothers’ and fathers’ income on children’s
nutritional status in Guatemala”, Social Sciences and Medicine, 37 (11), 1303–12.
Goetz, Anne Marie and Rina Sen Gupta (1996), “Gender, power, and control in
rural credit programs in Bangladesh”, World Development, 24 (1), 45–63.
Hossain, Mahabub (1988), “Credit for alleviation of rural poverty: institute research
report 65”, February, Washington, DC: International Food Policy Research.
Khandker, Shahidur R., Baqui Khalily and Zahed Kahn (1995), “Grameen Bank:
performance and sustainability”, World Bank Discussion Paper no. 306,
Washington, DC.
Mayoux, Linda (1999), “Questioning virtuous spirals: microfinance and women’s
empowerment in Africa”, Journal of International Development, 11 (7), 957–84.
Pitt, Mark (1999), “Reply to Jonathan Morduch’s: Does microfinance really help
the poor? New evidence from flagship programs in Bangladesh”, typescript,
Providence, RI: Department of Economics, Brown University.
Pitt, Mark and Shahidur Khandker (1998), “The impact of group-based credit pro-
grams on poor households in Bangladesh: does the gender of participants
matter?”, Journal of Political Economy, 106 (5), 958–96.
Rahman, Aminur (1999), “Microcredit initiatives for equitable and sustainable
development: who pays?” World Development, 26 (12) December.
Rahman, Aminur (2001), Women and Microcredit in Rural Bangladesh: An
Anthropological Study of Grameen Bank Lending, Boulder, CO: Westview Press.
Shultz, T. Paul (1990), “Testing the neoclassical model of family labor supply and
fertility”, Journal of Human Resources, 25 (4), 599–634.
Skoufias, Emmanuel (2001), “Is PROGRESA working? Summary of the results by
an evaluation by International Food Policy Research Institute (IFPRI)”, IFPRI
Food Consumption and Nutrition Division discussion paper no. 118,
Washington, DC.
Thomas, Duncan (1990), “Intrahousehold allocation: an inferential approach”,
Journal of Human Resources, 25 (4), 635–64.
Thomas, Duncan (1994), “Like father like son, or, like mother like daughter: parental
education and child health”, Journal of Human Resources, 29 (4), 950–88.
Udry, Christopher (1996), “Gender, agricultural production, and the theory of the
household”, Journal of Political Economy, 104 (5), 1010–46.
Index
A Little World 79 India 72, 73, 74
administrative costs 87, 90 and Internet 20
Africa 18, 20, 21, 113–14, 118 and mobile phone technology 13–14,
see also individual countries 20, 79, 80, 82, 83, 98
agencies see aid agencies; development prudential regulation 86, 94, 95, 96,
agencies; NGOs (non- 97–8, 105
governmental organizations) and securitization of micro-loans
agency costs 112 35
agricultural loan initiatives 4 supervision 98–100, 102, 103, 105
agricultural pricing information 13 women’s empowerment 18, 19, 20
agricultural production 112 see also bank failures; commercial
aid 111 banks; overbanking; rural
aid agencies 5, 92 banks; rural branches of
airtime credits 76, 77, 78 nationalized banks; village
Allen, Hugh 113 banks; individual banks
AML (anti-money laundering) 16–17, barcode-reading point-of-sale (POS)
77–8, 82, 89–90 terminals 12–13
Anderson, Siwan 111 Barth, J.R. 105
Asia 74 Beam 74, 80–83
see also Asia Pacific; East Asia; BlueOrchard 27, 28, 32, 46
South Asia; individual countries Bolivia 89, 91–2, 93–4
Asia Pacific 78 BOMSI (BlueOrchard Microfinance
see also individual countries Securities I) 27–32
asset class 40–41 bonds 28, 30, 38, 50, 53–5, 62–8
assets, median total of lending borrower characteristics’ information
institutions 20, 21 59, 69–70, 73, 90, 95
assortative matching 11–12, 19 borrowers 3, 7–8, 19, 26, 47, 48
see also borrower characteristics’
Baland, Jean-Marie 111 information; consumer
BancoSol 92 protection; men; self-help
Bangladesh 4, 100, 109–10, 113 groups (SHGs); solidarity
see also Bangladesh Bank; BRAC; groups; women; individual
BRAC securitization; Eastern borrowers
Bank Limited (EBL); Grameen BRAC 47–9
Bank; PKSF (Palli Karma BRAC securitization
Sahayak Foundation); Proshika currency risk 46
Bangladesh Bank 49–50 future prospects 69–70
bank failures 86, 93–4, 98, 105 performance 46, 61–8
banks political economy considerations 49–
and CDOs 31, 32 53
costs of access to 7 structure overview 35, 53–60
and equity 9, 37 transaction rationale 47–9
123
124 Index
branching requirements 96 m-commerce 78

branchless banking 97–8 micro-loans 87–8
see also mobile phone technology prudential regulation 87
Brazil 12–13, 111 women as microfinance customers
BRI (Bank Rakyat Indonesia) 94 112
broad borrowers 8 see also administrative costs; fees;
BTMs (biometric teller machines) 13 transaction costs
Burkina Faso 112 CRAB (Credit Rating Agency of
Bangladesh) 50, 53
capital adequacy 93–4 credit enhancements, BRAC
capital calls 101 securitization 55–6, 62–8
capital markets 5, 9–11 credit history 4, 8, 112
see also domestic capital markets; credit ratings 6, 11, 35, 38, 50, 62, 70, 89
international capital markets see also CRAB (Credit Rating
capital structure 8–9 Agency of Bangladesh);
Caprio, G. 105 MicroRate
cash 14, 75, 76, 77, 111, 113, 114, 115 credit rationing 3
cash flow 60–61, 62, 63, 65, 68, 70 credit supply 4
CDOs (collateralized debt obligations) credit unions 4, 7, 9, 18, 19, 20, 21, 94,
10, 27–34, 46 103
CFT (combating the financing of culture 20–21
terrorism) 17, 82, 89–90 see also social norms
CGAP (Consultative Group to Assist currency 33
the Poor) 6–7 see also currency risk; local currency
child health 111 currency risk 33, 46, 49–50
China 78
Citibank 53 data 56–60, 69–70
Citigroup 10 debt/equity ratio 9
co-signers as borrowers’ restrictions 95 default insurance 3
collateralized loans 94–5 default risk 11, 15, 19
commercial banks 26, 38, 39–40, 72, defaults, loan see loan defaults
74, 93–4 delegated supervision 103–4
commercial investors 29–31, 35–9 delinquent loans
Compartamos 6, 10, 37, 88 and BRAC securitization 55, 56, 57,
competition, commercial and non- 59–60, 61–2, 63–7, 68
commercial investors 38–9 collateralized versus uncollateralized
consumer protection 17, 88–9 94–5
contracts see loan contracts; puts demography, India 71–2
cooperatives 18, 20 deposit insurance 16, 97
see also credit unions; self-help deposits/savings
groups (SHGs); solidarity BRAC microfinance funding 48
groups in domestic capital markets 26
correspondent banking 12–13 India 71, 72, 74
corruption 77, 105 regulation 16, 86–7, 91, 92, 106
costs unsupervised small community-
access to credit and financial services based intermediaries 100–101
6–9, 14 and women 111, 115
cash handling 75, 76 development agencies 26, 27, 30, 35,
electronic matching of borrowers 38–9, 40, 101
and lenders 11 see also FMO; KfW
Index 125
Dexia Microcredit Fund 27 funding sources 48, 51

domestic capital markets 10, 26, 40 funds 27–9, 32–4, 38
Done Card 79
donor funding 48, 92 G-cash 14
DSRAs (debt service reserve accounts) GCMF (Global Commercial
56, 69, 70 Microfinance Facility) 38
DWM (Developing World Markets) gender see men; women; women-only
26, 27, 28, 29–30, 31, 38, 41 solidarity groups; women’s
disempowerment; women’s
East Asia 18, 20, 21 empowerment
see also individual countries geographic variation, women’s
Eastern Bank Limited (EBL) 47, 53 empowerment 18, 19–21, 117–18
eChoupals 13 Gonzales, A. 7–8, 9, 18, 19
economic downturns 31, 41 governments 5, 9, 48, 51–3
see also bank failures; sub-prime see also prudential regulation;
mortgage crisis supervision
economy, India 71–2 Grameen Bank 4, 20, 87, 108, 113
education 108–9, 110, 111, 114, 116, Grameen Trust Chiapas, AC (GTC)
117–18 114–15, 116, 117–19
see also financial education; literacy group lending see self-help groups
levels; primary education (SHGs); solidarity groups;
electronic matching of borrowers and women-only solidarity groups
lenders 11–12 GSMA 74–8
Engle, Patrice 111 Guatemala 111
entrepreneurship 4, 19, 110, 116, 118
equity 9, 28, 35–7 health 2, 22, 108–9, 110, 111, 114, 116,
Equity Bank (Kenya) 10, 37 117–18
equity capital 10 hedging 38, 39–40, 44, 55–6
Europe 35, 37, 38 high-end borrowers 8
see also FMO; KfW high-risk borrowers 3
exchange rates 50 HNWIs (high net worth individuals)
exits 36, 37 32, 38
household expenditure 110, 111, 114,
FATF (Financial Action Task Force) 117–18
89–90 household income 1, 108, 110, 111,
fees 34, 50 116, 117–18, 119
financial controls 42 housework 114–15
see also regulation; supervision
financial education 88–9, 101 ICICI 13, 46, 80
financial return 34, 36, 41–2 IFC Washington 10, 75–8
financial services 5, 7–9 IFIs (international financial
Fino 79–80 institutions) 10
FMO 35, 40, 49–50, 53 see also individual IFIs
food expenditures 110, 111, 114 illiquidity 34, 41
formal credit markets 2–3 income 1
see also banks see also household income
formal markets 2 India 4, 13, 15–16, 71–5, 78
fraud 77 see also ICICI; mobile phone
friction, marital 108, 113, 114, 115, technology in India
119 individual loans 8, 9, 18, 19
126 Index
individual responsibility 42, 74 IPOs (initial public offerings) 10, 36, 37

Indonesia 87, 94, 99–100 ITC (India Tobacco Company) 13
informal credit markets 3–4 ITZ Cash 79
informal markets 3
information see agricultural pricing JiGrahak 79
information; barcode-reading joint liability 3, 8, 74
point-of-sale (POS) terminals;
borrower characteristics’ Kenya 10, 77, 98, 111
information; data; eChoupals; KfW 10, 35, 50, 53
information asymmetry; Khandker, Shahidur 109–10
information availability; KIVA 12
information disclosure;
information sharing; Internet; land ownership 110
loan files; MIS (management Latin America 4, 18, 19, 20, 21, 36, 87
information systems); reporting; see also Compartamos; individual
transparency countries
information asymmetry 2–3, 114, 119 legal vehicles see SPVs (special-
information availability 3 purpose vehicles)
information disclosure 14–15, 88–9 Lending Club 12
information sharing 70 Levine, R. 105
insider lending 96 licensing 91–2, 99–100, 102
insiders, and equity 36 liquidations 28, 36
interest payment frequencies 3 liquidity 34, 40, 44, 62
interest rate caps 88 see also illiquidity
interest rates 4, 6–7, 11, 14–16, 19, 48– literacy levels 12, 101
9, 52, 87–8, 91 Little World, A 79
international capital markets loan contracts 3, 8, 9, 11, 35
amount of investment 25 loan defaults 15–16, 29, 42, 59, 89, 94–5
asset class, on the path to 40–41 see also default insurance; default
CDOs versus funds 32–4 risk; delinquent loans
commercial investors, growing loan files 42, 95, 101
participation in CDOs 30–32 loan loss provision 95
commercial investors, introducing to loan maturity 48
CDOs 29–30 see also short-term loan contracts
and concerns about microfinance loan repayments
41–3 and consumer protection 89
equity 35–7 frequencies 3, 48
from funds to CDO 27–9 and inclusion of men in women’s
importance 25–7 solidarity groups 114
and local currency 33, 38, 39–40 and m-commerce 14, 76, 77
non-commercial investors 26, 27, self-help groups (SHGs) in India 74
37–9 women 4, 112
and securitization of micro-loans loan size 19–20, 26
34–5 see also broad borrowers; high-end
international money transfer services borrowers; low-end borrowers;
81, 82, 83 small borrowers
Internet 11–12, 13, 20, 70, 79 local commercial banks 38, 39–40, 53
intimidation 89 see also individual banks
IPA (Innovations for Poverty Action) local currency 33, 38, 39–40, 46, 49–50
116–18 logistics 57–8
Index 127
long-term investment 27–9 India see mobile phone technology

low-end borrowers 7, 8 in India
low-risk borrowers 3 and m-commerce 75–8
mobile phone technology in India
m-commerce 75–8, 79 future and challenges 81–3
marriage rates 115 industry 79–80
Mayoux, Linda 113–14, 118 and microfinance 74–5, 78
MChek 79 products 80–81
men 108, 110, 111–13, 114–18, 119 usage 72
Mexico 111, 114–15, 116–19 money see AML (anti-money
MF Analytics 53, 56 laundering); cash; cash flow; CFT
MFBA (Microfinance Bank of (combating the financing of
Azerbaijan) 38 terrorism); currency; fees; loan
MFIs (microfinance institutions) repayments; loan size; money
borrowers, global number of 26 laundering
capital adequacy 93–4 money laundering 77
costs 87–8 see also AML (anti-money
India 72–4 laundering)
loan size 19–20, 26 Morgan Stanley 32
mission drift 42–3
ownership requirements 97 NBFIs (non-bank financial
regulation see regulation institutions) 5, 7, 9, 18, 19, 20, 21
size 100 see also credit unions; PACS
stock exchange listing 10 (primary agriculture credit
supervision 99–104 societies); ROSCAs (Rotating
sustainability 92 Savings and Credit
and women’s empowerment 18, 19, Associations); self-help groups
20–21 (SHGs); village banks
see also banks; NBFIs (non-bank NGOs (non-governmental
financial institutions); organizations)
individual MFIs and capital calls 101
MFRC (Micro Finance Regulatory and capital markets 10–11
Council) 89 costs of access to financial services
micro-credit backed securities 5 7, 9
micro-insurance 81, 82, 83 in evolution of microfinance 5, 9
micro-investment 10, 81, 82, 83 ownership requirements 97
see also international capital markets and women’s empowerment 18, 19,
micro-payments 82, 83 20, 21
microfinance, evolution 4–6, 9, 108 see also individual NGOs
Microfinance Securities XXEB 26, 31 non-commercial investors 26, 27, 30,
MicroRate 6, 38–9 32, 37–9
minimum capital requirements 93, 99, non-profit organizations 26, 33–4
100 see also aid agencies; credit unions;
MIS (management information development agencies; NGOs
systems) 51, 53, 56–8, 69, 70 (non-governmental
mission drift 42–3 organizations); PACS (primary
MIX data 6, 7–8, 9, 18–21, 87–8, 92 agriculture credit societies);
mobile phone technology ROSCAs (Rotating Savings and
and banks 13–14, 20, 79, 80, 82, 83, 98 Credit Associations); self-help
and BRAC securitization 70 groups (SHGs); village banks
128 Index
non-prudential regulation 17, 86, 87– prudential regulation

90, 103 adjusting regulations to fit products
and institutions 92–8
Omidyar, Pierre 5 critique 105–6
OPIC (Overseas Private Investors versus non-prudential regulation
Corporation) 30 86–7
OPORTUNIDADES 111 regulation as promotion 90–91
origination risk 35 regulation that follows the market
over-collateralization 63–8, 69–70 91–2
overbanking 33, 41–2 puts 36
ownership requirements 97 PWC (PriceWaterhouseCoopers) 53,
69
PACS (primary agriculture credit
societies) 72 RBI (Reserve Bank of India) 72
Pakistan 91 regulation
pawnbrokers 4 deposits/saving 16, 86–7, 91, 92, 106
pay-down structure 54–5 formal markets 2
payday lending 4 infrastructure problems 35
Paymate 79 interest rates 14–16
payment systems 74–5 and m-commerce 77–8
peer monitoring 14, 15, 18 and mobile phone technology in
peer pressure 15, 74, 112 India 82
performance, BRAC securitization 46, non-prudential regulation 17, 86,
61–8 87–90, 103
Philippines 14, 76–8, 98, 99–100 prudential regulation (see prudential
physical security 96 regulation)
Pitt, Mark 109–10 self-regulation 104
PKSF (Palli Karma Sahayak see also peer monitoring; supervision
Foundation) 48, 51, 53, 54 repayment capacity 89
political economy 49–53, 70, 92, 99, reporting 53, 56, 70, 96, 103, 104
104, 105 risk
pool maintenance, BRAC BRAC securitization 69, 70
securitization 53, 56 CDOs (collateralized debt
poor persons 2–3, 5, 22, 100–101, 110 obligations) 27–30, 31, 32, 33
see also poverty alleviation; poverty and costs of access to financial
measurement services 8–9
Portio Research 78 group lending and self-help groups
poverty alleviation 7–8, 110, 111 (SHGs) 3
poverty measurement 1 non-commercial investors as shield
prepayment risks 55–6, 60–61, 69 for commercial investors 37–8
prepayments, and BRAC securitization and overbanking 33, 41–2
55–6, 57, 60–61, 62–3, 64–7, 68, see also currency risk; default risk;
69 high-risk borrowers; low-risk
primary education 1, 22 borrowers; origination risk;
ProCredit Bank Bulgaria 35 prepayment risks; risk analysis;
ProCredit Holding AG 5, 35 risk aversion; risk/return ratios;
ProFund Internacional SA 36 security risks
promotion of microfinance 90–91 risk analysis 58–61, 70
Proshika 52–3 risk aversion 26, 70, 112
Prosper.com 11, 12 risk/return ratios 29–30
Index 129
ROSCAs (Rotating Savings and Credit social norms 112, 113–14, 118
Associations) 111 social pressure 15, 74, 112
RRBs (regional rural banks) 72, 73, 74 socially responsible investors see non-
rural areas 71, 72 commercial investors
see also RRBs (regional rural soft capital 5
banks); rural banks; rural solidarity groups 3, 8, 9, 14, 15, 18, 19
branches of nationalized banks; 21
village banks; village Internet see also women-only solidarity
rural banks 99–100 groups
see also RRBs (regional rural solvency 93–4
banks); rural branches of South Africa 13–14, 77, 89, 90, 91, 98
nationalized banks; village South Asia 18, 19, 20, 21
banks see also individual countries
rural branches of nationalized banks 5, SPVs (special-purpose vehicles) 28, 53,
7, 9, 72, 73 54, 55, 62–8
stand-alone loans see individual loans
sachet purchasing 76–7 Stephens, B. 10–11
savings see deposits/savings stock exchange listing 10
SCBs (scheduled commercial banks) stop-lending order 102
72, 74 sub-prime mortgage crisis 32
Schultz, Paul 111 substitution, and BRAC securitization
secondary markets 32, 43 55, 56, 63, 64–7, 68
Securities and Exchange Commission Superintendency of Banks (Bolivia)
(SEC) 50 89, 92
securitization of micro-loans 5, 10, supervision 96, 98–106
34–5, 46 sustainability of MFIs 92
see also BRAC securitization; CDOs Suvidha 74, 80, 81–3
(collateralized debt obligations) swaps 40
security risks 74, 90, 91, 100–101 SWIFT 74, 80, 81
see also physical security syndication 48, 49
self-employment 74
self-help groups (SHGs) 3, 14, 19, Tanzania 91
73–4 taxation 7, 82
see also solidarity groups; village technological innovation 5, 11–14,
banks; women-only solidarity 15–17
groups technology 2, 8, 22
self-regulation and supervision 104 see also Internet; mobile phone
servicers 35 technology; technological
see also BRAC securitization innovation
short-term investment 27, 50, 53, Thailand 111
54–5 Thomas, Duncan 111
short-term loan contracts 3, 11, 35 TIAA-CREF 5
Skoufias, Emmanuel 111 transaction costs 2, 7, 73, 78, 112
small borrowers 73, 87, 90, 91, transparency 30, 33
100–101 trustees 28, 30, 47, 53
see also low-end borrowers
small community-based intermediaries Udry, Christopher 112
100–101 uncollateralized loans 94–5
SMS (short message service) 78, 79, underclass 76
80–81 United Kingdom 11, 12, 111
130 Index
United States 4, 11, 12, 32 geographic variation 18, 19–21,

see also IFC Washington; 117–18
MicroRate; OPIC (Overseas and health 108, 110, 111, 114, 116,
Private Investors Corporation) 117–18
unsecured lending limits 94 and household expenditure 110, 111,
USAID 14 114, 117–18, 119
usury limits 87–8 and household income 108, 111, 116,
117–18
village banks 5, 8, 9, 18–19 and housework 114–15
see also rural banks; rural branches individual versus solidarity group
of nationalized banks lending 18–19, 21
village Internet 13 and lending institutions 18, 19,
violence 89 20–21, 108
volatility 41, 48–9 and level of aggregation 20–21
loan size 19–20
Wallet 365 79 and marital friction 108, 113, 114,
women 4, 111, 112, 115 115, 119
women-only solidarity groups 114–19 measurement 115–19
womens’ disempowerment 108–9, 113, microfinance in India 72
115–19 SHG (self-help group) lending in
women’s empowerment Asia 74
agricultural inputs and productivity and social norms 112, 113–14, 118
112 World Bank 1, 74, 110
assets 20, 21
and education 108, 110, 111, 114, Yanus, Muhammad 4, 113
116, 117–18
entrepreneurship 4, 19, 110, 116, 118 Zambia 92
in evolution of microfinance 4, 108 Zopa.com 11, 12

Microfinance

Caricato da

Informazioni sul documento

Descrizione originale:

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Microfinance

Caricato da

Copyright:

Formati disponibili

Microfinance

All rights reserved. No part of this publication may be reproduced, stored in

Edward Elgar Publishing, Inc.

A catalogue record for this book

Library of Congress Control Number: 2008935924

ISBN 978 1 84720 920 7

1 The changing landscape of microﬁnance 1

of developing countries. The generator was featured in Wired magazine in

supervision of microﬁnance, and teaches regularly on that topic at the

Brad Swanson is a partner in Developing World Markets, a socially respon-

FORMAL AND INFORMAL MARKETS

Capital markets and ﬁnancial institutions help provide mechanisms for

The properties of financial markets where there are significant differ-

Informal credit markets described in the previous section have preceded

Over the period 1980–90s many microﬁnance institutions began to

QUESTIONS ADDRESSED IN THIS BOOK AND A

What are the costs of accessing credit in microﬁnance? Does it depend on

Cost of Access to Credit

Lacking micro-level data, it is diﬃcult to estimate the costs of obtaining

Table 1.1 Costs of obtaining ﬁnancial services—variations across

Institution 2003 2004 2005

Source: Gonzalez (2007).

Table 1.2 Costs of obtaining ﬁnancial services—variations across target

Target Group 2003 2004 2005

Source: Gonzalez (2007).

Contracting and cost of access

Capital structure and default risk

Table 1.3 Costs of obtaining ﬁnancial services—variations across

Contracting Method 2003 2004 2005

Note: a. Group lending.

Source: Gonzalez (2007).

of the consumers of ﬁnancial services. We now turn to these issues. In the

Role of Capital Markets in Financing and Investment in Microﬁnance

Chapters 2 and 3 explore the role of capital markets in the ﬁeld of

There are other microﬁnance institutions that appear to rely on domes-

services and increased commercialization, with the median commercial

Technology and Microﬁnance—Opportunities and Challenges

In an earlier sub-section, we noted that the borrowing rates in microﬁnance

Electronic matching of borrowers and lenders: search and delivery costs

Table 1.4 Matching borrowers and lenders—some innovations

Organization Services Provided Some Attributes

Source: Author’s compilation.

Correspondent banking and biometric authorization of credit22

Village Internet centers and eChoupals

Mobile phone delivery of ﬁnancial services25

number of factors such as: (1) frequency of compounding and payments,

Defaults in 2006 on micro-loans in Andhra Pradesh27

Savings and intermediation

1. Should third parties (correspondents, for example) be permitted to

(AML) rules, as well as rules for combating the ﬁnancing of terrorism

Finally, issues of consumer protection also have to be addressed.

Gender Empowerment and its Characterization

● Is this empowerment related to the nature of the lending institution?

Table 1.5 Gender empowerment—percentage of women borrowers across

Region Banks Cooperatives and Non-bank NGOs

Source: MIX data of 2006 and author’s calculations.

Evidence on gender empowerment

Table 1.6 Percentage of women borrowers across diﬀerent institutions

Year Banks Credit Unions NBFI NGO

Source: Gonzalez (2007).

Table 1.7 Percentage of women borrowers across diﬀerent contracting

Year Individual Individual/Solidarity Solidarity Village Banking