sustainability of indian microfinance institutions ||

179
India Studies in Business and Economics Nadiya Marakkath Sustainability of Indian Microfinance Institutions A Mixed Methods Approach

Upload: nadiya

Post on 20-Dec-2016

227 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: Sustainability of Indian Microfinance Institutions ||

India Studies in Business and Economics

Nadiya Marakkath

Sustainability of Indian Micro� nance InstitutionsA Mixed Methods Approach

Page 2: Sustainability of Indian Microfinance Institutions ||

Sustainability of Indian Microfinance Institutions

Page 3: Sustainability of Indian Microfinance Institutions ||

India Studies in Business and Economics

For further volumes:http://www.springer.com/series/11234

Page 4: Sustainability of Indian Microfinance Institutions ||

Nadiya Marakkath

Sustainability of IndianMicrofinance Institutions

A Mixed Methods Approach

123

Page 5: Sustainability of Indian Microfinance Institutions ||

Nadiya MarakkathSchool of Management & Labour StudiesTata Institute of Social SciencesCentre for Social EntrepreneurshipMumbai, Maharashtra, India

ISBN 978-81-322-1628-5 ISBN 978-81-322-1629-2 (eBook)DOI 10.1007/978-81-322-1629-2Springer New Delhi Heidelberg New York Dordrecht London

© Springer India 2014This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part ofthe material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,broadcasting, reproduction on microfilms or in any other physical way, and transmission or informationstorage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodologynow known or hereafter developed. Exempted from this legal reservation are brief excerpts in connectionwith reviews or scholarly analysis or material supplied specifically for the purpose of being enteredand executed on a computer system, for exclusive use by the purchaser of the work. Duplication ofthis publication or parts thereof is permitted only under the provisions of the Copyright Law of thePublisher’s location, in its current version, and permission for use must always be obtained from Springer.Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violationsare liable to prosecution under the respective Copyright Law.The use of general descriptive names, registered names, trademarks, service marks, etc. in this publicationdoes not imply, even in the absence of a specific statement, that such names are exempt from the relevantprotective laws and regulations and therefore free for general use.While the advice and information in this book are believed to be true and accurate at the date ofpublication, neither the authors nor the editors nor the publisher can accept any legal responsibility forany errors or omissions that may be made. The publisher makes no warranty, express or implied, withrespect to the material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Page 6: Sustainability of Indian Microfinance Institutions ||

Foreword

As the world poured aid money into a bucket of Danaides to quench the thirst of thepoor, it searched for solutions which could help the poor raise themselves up withoutdependency. One insight came from entrepreneurial solutions, but the poor missedat least one vital lubricant for successful microenterprise: credit. Early experimentswith microcredit, in Brazil, Bangladesh and Bolivia, led to a glimmer of hope. Thesemicrocredit experiments were repeated and the hope turned into belief. They weresuccessfully scaled and replicated in other parts of the world, and the belief turnedinto confidence and expectation that they would work each time, every time, at leastmost of the times: the poor pay back.

The world was not perfect, and the small loan size led to high transaction cost.These high transaction costs were used to explain the relatively high interest ratesand the lack of profits. The search for efficiency in reducing operational costsstarted.

Donor money poured in to help scale up further. As some of the microfinanceinstitutions went beyond their break-even point, they realized that donor fundingcould now be replaced with commercial financing. Commercial financing renewedthe interest in the search for operational efficiency.

This expectation of profit led to investment, and the industry scaled up fast,very fast, and it became profitable, very profitable. Irrational exuberance set inand was communicated to donors, investors and the development community andthe investment turned into a fantasy that credit alone could solve poverty. Fromthis fantasy, mainstream lenders wondered that if microentrepreneurs could repayentrepreneurial microcredit, then surely microconsumers could repay consumermicrocredit. The gold rush of microcredit was on and the stampede could only endbadly. The fantasy led to a fall.

This fall came, ironically, not because the microfinance institutions becameunprofitable, but because they became too profitable. This high profitability fuelledtheir fantastic growth by raising their sustainable growth rate. The high profitabilitywas initially driven by high interest rates, but clearly as competition came in, themost efficient could grow faster and the reduction of operational costs becameimportant in a world where the fittest survived. The high profitability made people

v

Page 7: Sustainability of Indian Microfinance Institutions ||

vi Foreword

wonder who was gaining from microfinance: the poor or the creditors? Thedisillusion led to regulators stepping in to control the microcredit sector fromexploiting the poor. Today, questions are being raised on the ethics of microfinanceas to product, price, delivery and even its brand image.

The ethical debate will last a bit and certainly new deontological practices will beestablished, some by the industry and some by regulation. But once the debate hasdied down, we will come back to the main question: How can we reduce transactioncosts and increase efficiency so that the surplus of the microentrepreneurs and of theindustry can both be increased? It is to these questions that the remarkable work ofNadiya Marakkath is addressed.

Dijon, France Arvind Ashta1 May 2013

Page 8: Sustainability of Indian Microfinance Institutions ||

Preface

As a child, I was taught to believe that we are all inherently good and it is our innatenature ‘to do good’. As I grew up, my interactions with the external world largelyconformed to this childhood belief. I found that each one of us had a genuine desire‘to do good for all’ and have at some point of time in our lives mentally devisedplans to create an inclusive world. But for the vast majority of us, the conversionof these dreams into reality was mostly thwarted by thoughts that wrestle around‘sustainability’. Ultimately, it boils down to the double bottom-line dilemma: CanI do good and simultaneously earn a surplus out of it, so as to ensure perpetuityof goodness? Thus, to my observation (right from childhood till date) sustainabilityhas been a pertinent issue that plays a decisive role in determining the translation ofgood thoughts into good actions.

As I grew up with this realization, I developed an active research interest indecoding the mystery behind the sustainability of social enterprises. In this book,I document my understanding, gained mostly during my doctoral days, about thesustainability of one of the most celebrated forms of social enterprise that we havein our times—the microfinance institutions. Microfinance institutions, popularlytermed as MFIs, work towards creating a financially inclusive world, by providingfinancial intermediation services to the poor and hitherto excluded masses. Themajor challenge before these institutions is to attain operational self-sustainability(OSS), by earning enough self-generated revenue to cover its high intermediationcosts. Therefore, in this book, using a mixed-methods approach, I investigate thesustainability issues of Indian MFIs. What are the determinant and discriminantfactors for the sustainability of Indian MFIs? How are these factors being managedby operationally efficient MFIs in India, which remained sustainable at reasonableinterest rates before the onset of the crisis and interest rate ceilings in the Indianmicrofinance markets? What does the Indian microfinance crisis teach us aboutsustainability management and mismanagement? It is to these questions that Iattempt to find answers in this book. Overall, the results of this mixed-methodsinvestigation are expected to enable Indian MFIs to march towards the attainmentof efficiency and sustainability in their operations, without losing their focus on

vii

Page 9: Sustainability of Indian Microfinance Institutions ||

viii Preface

client welfare. Since the investigation uses a mixed-methods approach, it has thesynergistic value of both quantitative and qualitative analysis, making the results cutacross disciplines, ensuring methodological replication of this investigation feasiblefor any social venture facing sustainability challenges.

Conforming to the mixed-methods style of investigation, the book is structuredas follows: To begin with, this book presents a general introduction to microfinance(Chap. 1), specific literature review on MFI’s sustainability (Chap. 2) and anoverview of the mixed-methods research design used for investigating the sustain-ability issues (Chap. 3). The mixed-methods investigation pursued is then discussedas four specific objectives, each dealt with in separate chapters of the book. Chapter4 deals with the quantitative phase of the investigation, which aims to identify thefactors determining and discriminating the OSS of Indian MFIs. Chapter 5 dealswith the intermediate participant selection phase of the investigation, which aims toidentify efficient and sustainable Indian MFIs that can be benchmarked as referencegroups for other MFIs in the industry. Chapter 6 deals with the qualitative phaseof the investigation, which aims to document how efficient and sustainable MFIsare managing the factors determining and discriminating their OSS status. Chapter7 then goes to discuss about the dangers involved in the mismanagement of thesefactors by reflecting on the crisis that hit the Indian microfinance industry. Finally,the book concludes with Chap. 8 which presents a summary of the findings drawnfrom each of the three analytical phases pursued in this book. The implications ofthese findings are discussed as the key contributions made by this book.

Acknowledgements As mentioned at the onset, this book draws inspiration from the innategoodness that I have seen latent in every human being that I have interacted with. Therefore, Iconsider each of them my partner in bringing out this book. They are simply too numerous tomention individually, but I owe a deep and genuine sense of gratitude to each of them. In particular,however, I would like to thank my family (each and every one of them, to whom I owe everythingin life, that I proudly call today as ‘mine’), my alma mater (all those academic institutions which Iwas affiliated to—from KG to Ph.D.—and all those teachers out there from whom I picked up whatI consider ‘my way of life’), my mentors (but for whose timely intervention in my academic lifeI would consider myself ‘totally lost’), my research supervisors (all three of them, without whomI would never call myself a ‘researcher’), my research collaborators (the very many with whom Ilearnt the most intellectually stimulating process called ‘reflexivity’), my reviewers (without whosesharp observations I wouldn’t have appreciated the joy inherent in the process called ‘refinement’),my colleagues (all of them who ‘reinforced’ my belief: ‘Work is the most meaningful expressionof life!’), my students, potential social entrepreneurs (the co-learners, with whom I love to growevery day, without ever getting older) and my friends (who seemingly appear scattered across theglobe but are safely residing in my heart) for kindling and nurturing in me the childlike curiositycritical for any researcher for her investigation.

This research investigation wouldn’t have attained full fruition had it not been published anddisseminated as a book. I thank the entire team at Springer and SPi Technologies for makingthis happen. Lastly, I thank in advance every prospective reader of this book, whom I see as anembodiment of goodness. I earnestly wish this book could be of help to you in manifesting yourinnate goodness, with the requisite focus and flow on its sustainability.

Mumbai, India Nadiya Marakkath

Page 10: Sustainability of Indian Microfinance Institutions ||

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 An Overview of the Concept of Microfinance . . . . . . . . . . . . . . . . . . . . . . . 2

1.2.1 Definition of Microfinance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2.2 Definition of Microfinance Institutions. . . . . . . . . . . . . . . . . . . . . . . 21.2.3 Objectives of Microfinance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.3 An Overview of the Emergence of Microfinancein Global Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.4 An Overview of the Emergence of Microfinancein Indian Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61.4.1 Phase 1: Old Paradigm of Microfinance. . . . . . . . . . . . . . . . . . . . . . 61.4.2 Phase 2: New Paradigm of Microfinance. . . . . . . . . . . . . . . . . . . . . 10

1.5 An Overview of the Operational Self-Sustainabilityof Microfinance Institutions in India. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

1.6 Statement of Research Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141.7 Research Aim and Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141.8 Database, Sample and Methodology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151.9 Scope of the Research Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161.10 Organization of the Book. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

2 Literature Review: The Sustainability Debate . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212.1 Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212.2 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

2.2.1 Sustainability Definition and Metrics . . . . . . . . . . . . . . . . . . . . . . . . 212.2.2 Challenges Faced by Microfinance Institutions

in Attaining Sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.2.3 Relevance of Sustainability for Microfinance Institutions. . . 242.2.4 Balancing the Dual Goals of Microfinance Institutions . . . . . 25

2.3 Summary of Inferences and Research Gaps Existingin Microfinance Literature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

ix

Page 11: Sustainability of Indian Microfinance Institutions ||

x Contents

2.4 Research Objectives Revisited Along with AssociatedLiterature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302.4.1 Objective 1: To Identify the Factors Affecting

the Operational Self-Sustainability of IndianMicrofinance Institutions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

2.4.2 Objective 2: To Identify the FactorsDiscriminating the OperationalSelf-Sustainability Status of IndianMicrofinance Institutions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

2.4.3 Objective 3: To Identify Efficient and SustainableIndian Microfinance Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

2.4.4 Objective 4: To Understand Aboutthe Management of the Factors Affectingand Discriminating the OperationalSelf-Sustainability of Indian Microfinance Institutions . . . . . 32

2.5 Expected Value Additions from This Research Workto Microfinance Literature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

2.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

3 Research Objectives and Design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393.1 Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393.2 Research Objectives and the Sequential Explanatory

Mixed-Methods Research Design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393.3 Paradigm Elements in Mixed-Methods Research Design . . . . . . . . . . . 423.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

4 Quantitative Phase: Identification of Factors Affectingand Discriminating Sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454.1 Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454.2 Structure of the Quantitative Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454.3 Literature Review on the Factors Affecting

the Operational Self-Sustainability of Microfinance Institutions . . . . 464.3.1 Portfolio Risk Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474.3.2 Capital Structure Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474.3.3 Development Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474.3.4 Growth Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484.3.5 Institutional Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

4.4 Data, Theoretical Model and Hypotheses . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494.4.1 Dependent Variable. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514.4.2 Independent Variables and Hypotheses . . . . . . . . . . . . . . . . . . . . . . 51

4.5 Multiple Regression Analysis and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

Page 12: Sustainability of Indian Microfinance Institutions ||

Contents xi

4.6 Selection of Probable Discriminatorsfor the Discriminant Analysis Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594.6.1 Revenue Generation Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594.6.2 Cost-Efficiency Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

4.7 Data and Model for Discriminant Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 604.8 Multiple Discriminant Analysis Models and Results . . . . . . . . . . . . . . . . 614.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

5 Intermediate Participant Selection Phase: Assessmentof Efficiency and Sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 735.1 Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 735.2 Structure of the Intermediate Participant Selection Phase . . . . . . . . . . . 735.3 Literature Review on Microfinance Institution

Efficiency Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745.4 Data Envelopment Analysis Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . 755.5 Sample Data and Specification of Inputs and Outputs

for the Data Envelopment Analysis Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 785.6 Empirical Analysis and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

5.6.1 Efficiency Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815.6.2 Benchmarking Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845.6.3 Sustainability Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

5.7 Affect of Institutional Specific Factors on Efficientand Sustainable Microfinance Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

5.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

6 Qualitative Phase: Management of Factors Affectingand Discriminating Sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 976.1 Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 976.2 Overview of the Qualitative Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

6.2.1 Formulating the Interview Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 986.2.2 Pilot Testing the Interview Guide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996.2.3 Conducting the Final Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996.2.4 Preliminary Exploration of Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996.2.5 Data Analysis and Documentation of Strategies . . . . . . . . . . . . . 996.2.6 Triangulation Using Quantitative Data . . . . . . . . . . . . . . . . . . . . . . . 101

6.3 Discussion on the Relationship Shared by the FiveFactors with Operational Self-Sustainabilityof Microfinance Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1016.3.1 Portfolio Risk Factor: Mapping the Negative

Relationship Between Portfolio RiskGreater Than 30 Days and OperationalSelf-Sustainability Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Page 13: Sustainability of Indian Microfinance Institutions ||

xii Contents

6.3.2 Growth Factor: Mapping the PositiveRelationship Between Gross Loan Portfolioand Operational Self-Sustainability Ratio . . . . . . . . . . . . . . . . . . . . 102

6.3.3 Development Factor: Mapping the NegativeRelationship Between Average Loan Size PerBorrower and Operational Self-Sustainability Ratio . . . . . . . . 103

6.3.4 Institutional Factor: Mapping the NegativeRelationship Between Usage of Self-HelpGroup Credit Delivery Model and OperationalSelf-Sustainability Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

6.3.5 Cost-Efficiency Factor: Mappingthe Discriminatory Relationship Sharedby Operating Cost Per Borrower to OperationalSelf-Sustainability Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

6.4 Discussion on the Management of the Five Factors. . . . . . . . . . . . . . . . . . 1086.4.1 Portfolio Risk Factor: Strategies and Policy Suggestions . . . 1096.4.2 Growth Factor: Strategies and Policy Suggestions . . . . . . . . . . 1136.4.3 Development Factor: Strategies and Policy Suggestions . . . . 1186.4.4 Institutional Factor: Strategies and Policy Suggestions . . . . . 1216.4.5 Cost-Efficiency Factor: Strategies and Policy

Suggestions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

7 Qualitative Phase: Mismanagement of the FactorsAffecting and Discriminating Sustainability—Learningsfrom Indian Microfinance Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1357.1 Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1357.2 Indian Microfinance Crisis: A Learner’s Perspective . . . . . . . . . . . . . . . . 1357.3 Discussion on the Mismanagement of the Five Factors. . . . . . . . . . . . . . 136

7.3.1 Portfolio Risk Factor: Over-Indebtednessand Coercive Recovery Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

7.3.2 Growth Factor: Expansion Plans and Investor Pressures . . . . 1377.3.3 Development Factor: Multiple Borrowings

and Client Suicides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1387.3.4 Institutional Factor: Supply Driven Credit Model . . . . . . . . . . . 1387.3.5 Cost-Efficiency Factor: Strained Customer Relations . . . . . . . 139

7.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140

8 Summary of Findings, Implications and Conclusion . . . . . . . . . . . . . . . . . . . . 1418.1 Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1418.2 Summary of Findings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

8.2.1 Findings from Phase 1: Quantitative Phase . . . . . . . . . . . . . . . . . . 143

Page 14: Sustainability of Indian Microfinance Institutions ||

Contents xiii

8.2.2 Findings from Phase 2: Intermediate ParticipantSelection Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

8.2.3 Findings from Phase 3: Qualitative Phase. . . . . . . . . . . . . . . . . . . . 1448.3 Implications of the Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147

8.3.1 Practical Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1478.3.2 Theoretical Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1498.3.3 Policy Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

8.4 Limitations of the Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1528.4.1 Limitations of the Quantitative Phase . . . . . . . . . . . . . . . . . . . . . . . . 1528.4.2 Limitations of the Intermediate Participant

Selection Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1528.4.3 Limitations of the Qualitative Phase. . . . . . . . . . . . . . . . . . . . . . . . . . 153

8.5 Scope for Further Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1538.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155Appendix 1: Operational Definition and Distinction BetweenOperational Self-Sustainability Ratio, FinancialSelf-Sustainability Ratio and Subsidy Dependence Index. . . . . . . . . . . . . . . . . . . 155Appendix 2: Charnes, Cooper and Rhodes Model and Banker,Charnes and Cooper Model Formulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156Appendix 3: Interview Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158Appendix 4: Summary of the Qualitative Data CollectedDuring the Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160Appendix 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166

Page 15: Sustainability of Indian Microfinance Institutions ||
Page 16: Sustainability of Indian Microfinance Institutions ||

Abbreviations

BCC Banker, Charnes and CooperCCR Charnes, Cooper and RhodesCEO Chief Executive OfficerCGAP Consultative Group to Assist the PoorCRISIL Credit Rating and Information Services of India LimitedCRS Constant Returns to ScaleDEA Data Envelopment AnalysisDEAP Data Envelopment Analysis ProgramDRS Decreasing Returns to ScaleGNI Gross National IncomeIPO Initial Public OfferingIRS Increasing Returns to ScaleINR Indian RupeeIT Information TechnologyMFI Microfinance InstitutionMIS Management Information SystemMIX Microfinance Information ExchangeM-CRIL Microfinance Credit Rating Information LimitedMPSS Most Productive Scale SizeNABARD National Bank for Agriculture and Rural DevelopmentNBFC Non-Banking Finance CompanyNCAER National Council for Applied Economic ResearchNGO Non-Governmental OrganizationNIRS Non-Increasing Returns to ScaleNPAs Non-Performing AssetsOSS Operational Self-SustainabilityPDA Personal Digital AssistantPOS Point of SalePTE Pure Technical EfficiencyRBI Reserve Bank of IndiaRFAS Rural Finance Access Survey

xv

Page 17: Sustainability of Indian Microfinance Institutions ||

xvi Abbreviations

RRB Regional Rural BanksSBI State Bank of IndiaSE Scale EfficiencySEWA Self-Employed Women’s AssociationSHG Self-Help GroupSIDBI Small Industries Development Bank of IndiaSKS Swayam Krishi SangamSPSS Statistical Package for Social ScienceTE Technical EfficiencyUSD United States DollarsVRS Variable Returns to Scale

Page 18: Sustainability of Indian Microfinance Institutions ||

Chapter 1Introduction

1.1 Preface

Microfinance refers to the provision of financial services to low-income clients. Byproviding financial access to the poor clients, microfinance plays a decisive roletowards financial inclusion. It economically empowers the poor and integrates themto the mainstreams of the economy. The institutions that provide such financialservices to the poor are called microfinance institutions (MFIs). These MFIs act inan environment of high information asymmetric credit market risk, where there is adearth of information about the credit history of the poor clients. These informationasymmetric credit market risks are mitigated by the MFIs by using unconventionalgroup-lending models that work on joint-liability principle, sans collaterals. Thoughthis unconventional group-lending model has the potential to mitigate risk andfacilitate financial intermediation at the bottom of the pyramid, it has one majorchallenge associated with it—high intermediation costs. To cover these costs bygenerating a surplus from its operations and remain operationally self-sustainableare a formidable task for MFIs. Considering this operational challenge and thepertinence of sustainability as a means to an MFI’s social goal of poverty alleviation,institutionalist MFI practitioners and researchers advocate the practice of chargingcost-covering interest rates in the microfinance industry. This practice aids an MFIto attain an operationally self-sustainable status, but it has an inherent dangerassociated with. The danger is that cost-covering interest rate charged by MFIs canbe considered to be unreasonable if the costs of the MFIs are excessively high dueto inefficiencies in its operating structure. Practices of charging unreasonably highinterest rates have resulted in vulnerabilities in several microfinance markets, as itis tantamount to client exploitation. It was one of the reasons attributed to havedoomed a crisis in India—the world’s largest microfinance market—in the year2010. When the crisis hits the sector, the regulators intervened in the microfinancemarkets and fixed a reasonable interest rate as the ceiling for MFI operations. Asthe industry is recovering from the adverse effects of this crisis and when there arestill contentions as to whether the rate fixed by the regulator is good enough for the

N. Marakkath, Sustainability of Indian Microfinance Institutions:A Mixed Methods Approach, India Studies in Business and Economics,DOI 10.1007/978-81-322-1629-2__1, © Springer India 2014

1

Page 19: Sustainability of Indian Microfinance Institutions ||

2 1 Introduction

sustenance of the MFIs, through a research effort pursued in this book, the authorinvestigates about the issues related to the sustainability of Indian MFIs, using amixed-methods approach.

1.2 An Overview of the Concept of Microfinance

Before beginning with the research investigation, this section presents a genericintroduction to the concept of microfinance, the definition of MFIs and theobjectives of microfinance.

1.2.1 Definition of Microfinance

‘Microfinance refers to the provision of financial services to low-income poor andvery poor self-employed people’ (Otero 2000). According to Robinson (2001),microfinance refers to ‘small-scale financial services–primarily credit and savings–provided to people who farm or fish or herd; who operate small enterprises ormicroenterprises where goods are produced, recycled, repaired, or sold; who provideservices; who work for wages or commissions; who gain income from renting outsmall amounts of land, vehicles, draft animals, or machinery and tools; and toother individuals and groups at the local levels of developing countries, both ruraland urban’. Schreiner and Colombet (2001) define microfinance as ‘the attempt toimprove access to small deposits and small loans for poor households neglected bybanks’. Oikocredit (2005) regards microfinance to be a wide concept that has micro-credit as one component. It also has provision of additional non-credit financialservices such as savings, insurance, pensions and payment services, as its othercomponents.

1.2.2 Definition of Microfinance Institutions

Morduch (1999) defines MFIs as specialized financial institutions, united underthe banner of microfinance, sharing the commitment to work towards financialinclusion. According to Asian Development Bank, MFIs are defined as institu-tions whose major business is the provision of microfinance services, such asdeposits, loans, payment services, money transfers and insurance, to poor and low-income households and their microenterprises. Ledgerwood (1999) regards MFIs asproviders of such financial services to poor—mainly credit and savings—althoughinsurance and other payment services are rendered by some.

Sriram and Upadhyayula (2004) clarify the concept of MFIs, by negating certainplayers from the purview of MFI’s definition. A commercial bank downscaling its

Page 20: Sustainability of Indian Microfinance Institutions ||

1.2 An Overview of the Concept of Microfinance 3

operations to reach the poor and a moneylender catering to financial needs of thepoor are excluded from the definition, as they do not fulfil the value attributes asso-ciated with an MFI. As per the value attributes of an MFI, only if an institution hasdevelopmental roots and a non-exploitative intent in predominantly serving the poor,can it be termed as an MFI. Thus, as per this definition, even a non-governmentalorganization (NGO) can be regarded as MFI, only if it does microfinance as a coreactivity or has a separate division created to handle microfinance operations.

1.2.3 Objectives of Microfinance

As microfinance envisions providing financial access to the poor in a sustainablemanner, its main objectives are as follows:

(a) Financial inclusion: MFIs aim to provide financial access to the poor andlow-income category, who are hitherto unreached by the traditional banks,on account of the high credit risks associated with them. Thus, it aims tocreate a well-functioning financial system that is more inclusive in nature(United Nations Capital Development Fund 2006). According to Sen (1999),a well-functioning market system has the potential to confer freedom ofchoice to the poor and to remove their sources of their unfreedom, throughbroadening of their choice or reduction of their deprivation, itself resulting intheir development.

(b) Poverty reduction: By providing financial access to the poor and the weakersections of the society, MFIs address a major constraint faced by them:shortage of material capital, i.e. the input necessary to generate income (Otero2000). Thus, MFIs enable the poor to create wealth, deal with risks, developtheir microenterprises and smoothen their consumption patterns. All these areexpected to reduce their vulnerability in facing the calamities of life and toimprove their quality of life (Johnson and Rogaly 1997; Gibbons and Meehan2002; Simanowitz 2002; Armendariz de Aghion and Morduch 2005; Bakhtiari2006).

(c) Women empowerment: Since women are relatively more disadvantaged thanmen in financial matters, microfinance aims to empower women by providingthem financial access. By enabling women to secure access to financial sourcesand contribute to their family income, microfinance equips them to gain moreindependence and confidence in running their family. This objective of womenempowerment complements its aim of poverty alleviation, due to the fact thatwomen tend to spend more of their increased income on their households,children’s education and the family’s welfare than men (United Nations Fundfor Women 2001).

(d) Sustainability: Microfinance differs from other poor-financing efforts, in that itaims to alleviate poverty while paying for itself and perhaps even turning a profit(Brau and Woller 2004; Robinson 2001). Though the animating motivation

Page 21: Sustainability of Indian Microfinance Institutions ||

4 1 Introduction

behind the microfinance movement is ‘poverty alleviation through the financialinclusion’, the industry emphasizes the need for its players to be financiallysustainable. Rhyne (1998) remarks that sustainability is not an end in itself buta means for creating improved welfare, through financial outreach to poor.

1.3 An Overview of the Emergence of Microfinancein Global Scenario

Worldwide, accolades have been awarded to microfinance industry, since ProfessorMuhammad Yunus received the Nobel Peace Prize for his work that led to theformulation of an unconventional group-lending microcredit delivery model. Thisled to the recognition of the role that microfinance plays in poverty alleviationthrough financial inclusion, by the Nobel Prize Committee, in the year 2006. But theconcept of microfinance existed prior to this acknowledgement. It existed as grass-root movement in development realm before burgeoning as a global industry amongthe financial circles (Christen et al. 1995). This gradual transition as portrayedby Robinson (2001) and Consultative Group to Assist the Poor (CGAP 2006) isdiscussed below.

Early traces of microfinance can be related back to the informal savings andcredit groups that have operated for the poor, centuries ago. This include the‘susus’ of Ghana, ‘chit funds’ in India, ‘tandas’ in Mexico, ‘arisan’ in Indonesia,‘cheetu’ in Sri Lanka, ‘tontines’ in West Africa and ‘pasanaku’ in Bolivia. In 1700s,the Irish author Jonathan Swift initiated the earliest form of modern MFIs—theIrish loan fund system. The Irish loan fund system was designed to provide smalluncollateralized loans to rural poor. In 1800, various other formal institutions beganto emerge in Europe in the forms of people’s banks, credit unions and savings andcredit co-operatives. Of these, the credit unions developed by Friedrich WilhelmRaiffeisen gained wide acclaim in Europe and other North American States, inrelieving the rural poor from the clutches of usurious moneylenders. In 1895people’s banks became popular in Indonesia, and in 1900 the idea spread to LatinAmerica.

Between the 1950s and 1970s, supply-led government interventions, throughcommercial banks, co-operatives and rural development banks, were expedited. Thetargeted credits lent through these institutions were seldom successful due to thehigh information asymmetric credit market risk associated with microfinance.

Owing in part to these risk characteristics, traditional commercial banks werereluctant to operate in the microfinance segment. To induce financial intermediationat the bottom of the pyramid, governments assisted these banks by providingsubsidized funds for on-lending to the microfinance market. These subsidizedfunds were lent at below market interest rates, creating a negative spread inthe microfinance operations of the banks. These subsidized microfinance modelsresulted in mistargeting and debt waivers, working against the market dynamics.

Page 22: Sustainability of Indian Microfinance Institutions ||

1.3 An Overview of the Emergence of Microfinance in Global Scenario 5

2.SubsidizedInterest Rates

3. NegativeSpread

4. Mistargeting

5. Debt WaiversPoor Recovery

Rates

1. Subsidy

Fig. 1.1 Author’s schematicof subsidy trap (Source:Ledgerwood 1999)

Thus, traditional commercial banks operating in microfinance market werecaught in a subsidy trap, resulting in mounting losses for the lender (Yaron et al.1997). This as elaborated by Ledgerwood (1999) is depicted graphically in Fig. 1.1.

Against the backdrop of this market failure for the poor, separate financialinstitutions called MFIs emerged as a new paradigm in the annals of microfinancereforms. Unlike the traditional banks which used individualized banking model,these MFIs mitigated the information asymmetric credit market risks by usinggroup-based credit delivery models.

These MFIs began to provide finance to the poor clients organized as groups.Repayment rates were seen to be high on these group loans lent to poor clients, as thegroup members shared similar socio-economic characteristics. These MFIs emergedas a result of the experimental research undertaken in lines of group lending inBangladesh during the 1970s. In 1976 Professor Muhammad Yunus of Bangladesh,with his graduate students in Chittagong University, conducted an action-orientedresearch to disburse microfinance to the poor through an unconventional group-lending model. Under this model, the poor were organized as groups and were madejointly liable for the loans they received without any collateral backing. This modellater became popularized as the Grameen Bank concept of Bangladesh, which wonthe Nobel Prize in the year 2006.

Thus, though the award-winning concept of Grammen Bank was not the firstform of MFI, it definitely made the idea of microfinance popular. It led to therevolutionary shift from the supply-led government interventions to the financialsystem approach, where credit was viewed as a freely priced service that requiredless subsidization and rationing (Robinson 2001). It overturned the established ideasof the poor as beneficiaries of subsidized financial services and shattered stereotypesof the poor as not bankable (Mutua et al. 1996). It proved that through group-lending models that work on joint-liability principle, high repayment rates couldbe fetched from the poor, as all the members of the group shared the liability for the

Page 23: Sustainability of Indian Microfinance Institutions ||

6 1 Introduction

loans taken. It depicted that peer pressure and threat of social punishment within thegroups could effectively be used to replace the need for physical collateral in thesemodels. Thus, the poor were proved bankable by the group-lending microfinancemodels (Ghatak and Guinnane 1999). Robinson (2001) regards these microfinancepractices, built around providing uncollateralized small loans and acceptance of tinysaving deposits, as nothing short of a revolution or a paradigm shift in the world ofdevelopment.

1.4 An Overview of the Emergence of Microfinancein Indian Context

Sriram and Upadhyayula (2004), in their work, ‘The Transformation of the Microfi-nance Sector in India: Experiences, Options, and Future’ narrates the transformationexperiences of Indian NGOs into MFIs and contrasts this with internationalexperiences. While the Indonesian experience has been that of banks adopting MFImethods to mainstream financial services to the poor, the Bangladeshi experiencewas seen to be the transformation of a project (Prof. Mohammed Yunus’ action-oriented research on credit delivery to the poor) into an MFI. The Bolivianexperience was that of NGOs transforming to MFI-banks, and it shared somecommon notes with the Indian story of NGO-MFI evolution.

Before this NGO-MFI evolution in India is discussed, the series of experimentalmicrofinance reforms that has gone into its making is reviewed. The saga of Indianmicrofinance reforms is surveyed in this chapter by categorizing them into twophases. The two-phased approach used in the study is not a rigid demarcation basedon time periods. It is based on the competing thinking frames1 that underpin theparadigm shift seen in Indian microfinance initiatives, from government-subsidizedcredit delivery mechanisms to the specialized private financial intermediaries calledMFIs. The competing thinking frames that underpin the paradigm shift in thesephases are discussed in Table 1.1. Against the backdrop of these competingthinking frames, the study goes on to present a state-of-art analysis of the twophases/paradigms in Indian microfinance.

1.4.1 Phase 1: Old Paradigm of Microfinance

Right from the days of independence, provision of institutionalized microcreditto the rural poor gained prime importance in Indian microfinance policy reforms.

1Competing thinking frames are mental structures that force people to view the same objectivescenario from particular, limited and diverse perspectives, which is often coloured by theirsubjective beliefs. In this study, we portray how the concept of microfinance was viewed withsuch competing thinking frames.

Page 24: Sustainability of Indian Microfinance Institutions ||

1.4 An Overview of the Emergence of Microfinance in Indian Context 7

Table 1.1 Competing thinking frames

Phase I Phase IICompetingthinking frames Old paradigm of microfinance New paradigm of microfinance

Perception aboutmicrofinance andits sustainability

Microfinance aims at the provisionof subsidized microcredit to thepoor. The poor are unbankable,and therefore, microfinancecannot be rendered in aneconomically profitablemanner. Continued externalsupport in the form ofgovernment subsidies is neededto render these services in asustainable manner

Microfinance aims at the provision ofsustained financial services to thepoor—credit, saving, insuranceand other payment services. Thepoor are bankable, and therefore,microfinance can be rendered in aneconomically profitable manner.Subsidies are needed to supportmicrofinance activities; but in thedue course with self-generatedrevenues, these services can berendered in a self-sustainablemanner

Perception about thepoor clients

Poor clients are treated asbeneficiaries, who availsubsidized microcredit tomanage their financial needs,without much potential forsavings

Poor clients are treated as potentialcustomers, with savings potential.The poor demand a range offinancial services, to be providedon a sustained basis, in order tomanage their financial needs

Perception aboutrisk mitigationand collaterals inmicrofinance

Presence of high informationasymmetric risks that cannot bemitigated by the individualcredit delivery models justifiesthe use of collaterals inmicrofinance

Presence of high informationasymmetric risks that can bemitigated by group credit deliverymodels justifies the practice oftrust-based uncollateralizedmicrofinance

Perception aboutmicrofinanceservice delivery

Microfinance services are bestrendered through formal andsemi-formal banks in the publicsector, i.e. via commercialbanks, regional rural banks(RRBs) and co-operativesocieties

Microfinance can be renderedeffectively through private sectorMFIs, NGOs and Self-HelpGroups (SHGs)

With the unorganized sector dominating the microcredit scene,2 the presence of aformal institutionalized structure in microfinance was felt imperative. Based on thisobservation, the State Bank of India (SBI) was set up in 1955, 14 large commercialbanks were nationalized in 1969, and the National Bank for Agriculture and RuralDevelopment (NABARD) was created in 1982. Cooperative banks and regionalrural banks (RRBs) were set up, during the period 1950–1976, primarily to meetthe agricultural credit needs. In early 1980s six more banks were nationalized andmore branch expansions were undertaken (Thorat 2006; Leeladhar 2007). Despite

2All-India Rural Credit Survey (1954) conducted by Reserve Bank of India (RBI) portrays Indianrural poor to be dependent on local moneylenders, for more than 90 % of their financing needs.

Page 25: Sustainability of Indian Microfinance Institutions ||

8 1 Introduction

these efforts put in by successive governments to expand the reach of these formalfinancial intermediaries, the indented results did not materialize. Poor continued toaccess services from non-institutional sources to meet their financial exigencies.

The plethora of reforms3 based on subsidized microcredit was inherently inca-pable of providing sustainable financial services to the poor. Subsidized microcreditdelivery could provide only limited volume of cheap loans, which often ended upbeing allocated to the local elite, who are more influential in bypassing the deservingpoor. In addition to the misallocation, it resulted in high arrears4 and losses tothe service providers and funding agencies, depressing the viability of pro-poorfinancing programmes, in general, and sustainability of microcredit ventures, inparticular (Robinson 1995).

In the early 1990s, at the outset of the structural reforms, the microfinancingprofile remained far from satisfactory (Thorat 2006). The financial sector reformsthat ensued and swept a wave of competition and deregulation in the banking sector,as a whole (Narasimham 1991). With respect to microcredit, it did not bring muchof a change in the competitive scenario5 among the service providers. But it didset a new trend of autonomy in pro-poor financing. It liberalized the interest rates6

for cooperatives and RRBs, relaxed the controls on poor financing, reworked thesubheads under the priority sector credit and introduced prudential lending norms.

The reforms also initiated efforts to restructure and refinance the financiallydeteriorated RRBs. But a thorough overhaul of the flaws created by decadesof bureaucratic slouch, mismanagement and distorted incentives systems in theco-operative credit institutions and RRBs required much more than refinancing.Recapitalization, without concomitant reforms in its operational design, oftenculminated in postponing failure, hampering the sustainability of these institutions(Vaidyanathan 2004).

The post-liberalization scene resulted in mounting over dues for the banks. As onMarch 2004, priority sector lending constituted 47.5 % of the total nonperformingassets (NPAs) of the public sector banks. Provisioning norms for NPAs, furtherdeteriorated the banks, in the form of capital erosion. Negative spread on account

3Policy initiatives like directed credit programmes, subsidized interest rates, priority sector lending,lead bank scheme and service area approach sufficed mainly to comply with the quantitative targetsof poor lending, much needed for projecting the social face of banking. See Sinha and Patole (2003)for details.4Repayment rate of Integrated Rural Development Programme (IRDP), which delivered subsidizedcredit to 5.38 core families through commercial banks, was as seen to be as low as 25–35 %.Though it was considered to be the world’s largest microcredit programme, it is reported to haveresulted in huge losses for commercial banks. See RBI (1995) for more details.5Old paradigm was rooted in the belief that microfinance is best rendered via nationalization ofbanks than privatization. See Raj (1974) for more details.6Deregulation of interest rates was an integral part of financial sector reforms, intended to ensureefficiency allocation of resources and better price discovery. As per this mandate, apart frominterest rates on savings deposits and NRI deposits and export credit and small loans up to Rs.2 lakh, all other interest rates have been made flexible. See Narasimham (1991) for details.

Page 26: Sustainability of Indian Microfinance Institutions ||

1.4 An Overview of the Emergence of Microfinance in Indian Context 9

of mismatch between lending rates and deposit rates proved to be a disincentive forsavings mobilization. The loss on account of mismatch between lending rates andcost of financing resulted in a fall of nearly 30 % of net profits for the banks (Basu2008). Thus, in the old paradigm, banks internalized the notion that subsidizedmicrocredit lending is a loss-making activity, to be undertaken as a mandatory socialbanking practice.

In the old paradigm, the poor continued to resort to moneylenders for theirfinancial needs. Agrawal (2008) observes that even in cases where market supplyof low-cost microcredit was augmented through policy interventions in the oldparadigm, the poor preferred not to meet their entire financial demand from theformal financial sectors. Agarwal explains this anomaly by citing the price andnon-price barriers associated with microcredit delivery. The non-price barrierslike elaborate documentation and income assessments make the poor reluctant toapproach the formal banking system. To by-pass these non-price barriers and toeconomize their daily financial needs, they seek the assistance of local moneylen-ders than the formal financial institutions. The behavioural justification given forthe phenomenon is that financial exclusion is highly correlated with the social andself-exclusion traits, latent in the poor psyche (Sinclair 2001). Rural Finance AccessSurvey (RFAS 2003 as cited in Basu 2008) further goes on to quantify the non-pricebarriers, like the high rates of bribes to be paid and extended time taken for accessingfinance from formal sectors. Extracts from RFAS (2003) indicate that average briberate for availing a loan from commercial bank comes to nearly 10 % of the loanamount and the documentation time is close to 33 weeks, with high loan rejectionrates.

All this made financial access difficult under the old paradigm of microfinance,justifying the poor’s act of approaching the usurious moneylenders. In the financialdiaries of Ruthven (2001),7 one of his respondents gave a thought-provokingrationale for choosing non-institutional instruments, even if it meant paying usuriousrates. When I go to a money lender, it’s between him and me. I give my relatives noreason to talk (Ruthven 2001, p. 14). The message is thus clear—the convenience,speed and dignity conferred by close social circles in a locality reduce the non-pricebarriers and transaction costs for the poor. The old paradigm of microfinance wasunable to confer these benefits.

Despite these shortcomings, the positive impact of the policy efforts taken inthe old paradigm of microfinance was that it resulted in a huge increase in thebanking branch outreach in India, as the average population covered by a branchfell from 64,000 to 13,711 (Thorat 2007). This served as a good infrastructural basefor the microfinance practices, subsequently undertaken in the new paradigm ofmicrofinance.

7Ruthven (2001) used financial diaries to comprehend the financial instruments preferred by theIndian poor.

Page 27: Sustainability of Indian Microfinance Institutions ||

10 1 Introduction

1.4.2 Phase 2: New Paradigm of Microfinance

Indian Financial System did a lot of experimentations in the field of microfinance,before its transition to the new paradigm of microfinance. Lessons learnt about thenon-price barriers in old microfinance paradigm and introspective assessments ofinformal group credit-lending methods like chit-funds devised by the poor graduallyled to this transition.

Microfinance in the new paradigm had its modest beginning as a grass-rootdevelopment movement among NGOs in the early 1970s. The Self-Help Groups(SHGs), formed by these NGOs, were affinity groups of around 15–20 poorindividuals, mostly women with a homogeneous socio-economic background,sharing the willingness to improve their living conditions. The group membersprovided financial support to one another through internal credit assistance madefrom their pooled savings. This was an informal credit-lending method designedby the poor themselves to meet their consumption and productive needs. Afterinculcating financial discipline among themselves, these SHGs, formed under theaegis of the NGOs, persuaded the government to link themselves to formal financialinstitutions for sourcing additional funds and depositing their pooled savings.This, when acceded to, paved way for the India’s celebrated SHG-Bank LinkageProgramme.8 Later, some of the Indian NGOs, instead of merely performing the roleof a facilitator or promoter for microcredit, transformed themselves into specializedfinancial intermediaries called MFIs, constituting a niche industry with high growthpotentials. The Indian MFIs assumed heterogeneous forms comprising of non-banking finance companies (NBFCs), societies, trusts and co-operatives. Theyorganized the poor into groups and catered to their financial needs, instead of linkingthem to the banks. In 1973, an MFI called Self Employed Women’s Association(SEWA) was registered as a trade union in the district of Gujarat in India, to meetthe financial needs of bottom of the population pyramid. This institution called theMahila SEWA Co-operative Bank was the first MFI in India.

Thus, gradually what lied dormant as an informal lending method among thepoor eventually initiated a new paradigm in Indian microfinance reforms. The MFIsovercame the non-price barriers experienced in the old paradigm of microfinance asit provided speedy financial access to the poor, within their local socio-economiccircles using group-lending credit delivery models. Since the non-price barriersare low in a MFI model, the National Council for Applied Economic Research(NCAER 2011) study on small borrowings in India observes the over-all cost for

8In the 1991 RBI Circular, an announcement for linkage of informal SHGs, with the existingbanking system was made. In the following year NABARD launched a pilot project which linked500 SHGs with commercial banks. These incidents formally marked the advent of the newparadigm of microfinance. Banks were permitted to classify such microfinance lending under itsadvances to weaker sections under the priority sector lending norms.

Page 28: Sustainability of Indian Microfinance Institutions ||

1.4 An Overview of the Emergence of Microfinance in Indian Context 11

borrowings for the poor from an MFI to be least when compared to that of the costsassociated with formal financial institutions, SHG-Bank Linkage Programmes andother informal sources.9 This can be attributed mainly to the credit delivery modelused by MFIs which operates at the grass-root level.

The MFIs in India used two forms of group-lending credit delivery models—the home-grown SHG model (discussed earlier in this section) and the BangladeshiGrameen model for its credit delivery. Unlike the SHG model where groups areinitiated by the poor themselves, in the Grameen model, small-sized groups ofindividuals are formed among the poor by the MFIs for the purpose of microfinancedelivery. Weekly meetings are organized among the group members, and savingis made mandatory for them. Credit is not given to all members simultaneously,but all hope to have their turn and all stand jointly liable for amount lent to eachperson in the group. The principle of joint liability is used as collateral for theloans they receive from the MFI (Ghatak 2000). Repayment rate and efficiency wasseen higher under joint-liability contracts as compared to conventional individual-liability contracts because the former exploits a useful resource that the latterdoes not—the information that borrowers have about each other. Thus, repaymentrates close to 90 % was reaped by these NGO-MFIs in the new paradigm ofmicrofinance, which proved that lending to the poor is not a loss-making business(Thorat 2006).

Later in the early 2000s, the NGO-MFIs began to establish their permanentfooting in Indian financial system by transforming its informal nonprofit legal statusto formal pro-profit NBFC status. The snapshot of the share of poor clients servedby the formal and informal MFIs in India is portrayed by Srnec (2007) as givenbelow in Fig. 1.2.

As per NABARD’s statistics, after taking into account both these formal andinformal MFI in India, there are close to 800–1,000 MFIs operating in India, ason April 2011. Out of these 1,000 odd MFIs, nearly 52 MFIs are regulated NBFC-MFIs, and the rest are NGO-MFIs in the form of trusts, co-operatives and societies.As shown in Fig. 1.2, together they serve close to 70 million poor and financiallyexcluded masses in India as on April 2010. Sustaining the operations of these MFIsis crucial as the 70 million served by the MFIs constitute only a fraction of the700 million Indian people who lack access to essential financial services like credit,insurance and savings facilities and therefore constitute the potential target clientbase for microfinance services (Intellecap 2010).

9The overall cost of borrowing for the poor includes both the interest rate charged on loans andthe cost of non-price barriers associated with obtaining a loan from a given source. The cost ofnon-price barriers included wage loss due to time spent in getting the loan approved (opportunitycost), cost of travel, money spent on food while travelling to the source of loan, charges paidfor preparation of documents, additional charges (like stamp duty), payment of bribes and othercharges associated with insurance.

Page 29: Sustainability of Indian Microfinance Institutions ||

12 1 Introduction

Fig. 1.2 Comparison of formal and informal microfinance institutions in India (Source: Srnec2007)

1.5 An Overview of the Operational Self-Sustainabilityof Microfinance Institutions in India

Sustainability of an MFI in its primal sense is denoted by its operational self-sustainability (OSS). OSS denotes the ability of MFI to earn revenue to cover itscosts and reach the poor now and in the future (Schreiner 1996). More specifically,it is the ability of MFI to generate enough revenue from its operations to cover itsfinancing costs, transaction cost and loan loss provisions. It is captured by the OSSratio, which is the operating income10 of an MFI over the total costs of an MFI(i.e. operating costs11 C financing costs12 C loan loss provision13). A ratio above100 % denotes that MFI has enough operating income to cover its costs, indicatingan operationally self-sustainable status. Attaining OSS is imperative for the MFIto perpetually operate in the sector. But owing to the high information asymmetriccredit market risks in the microfinance market, the cost of intermediation is veryhigh making it difficult for MFIs to attain sustainability.14 In India, the industryaverage of costs ranges from cost of funds, nearly 10–14 %; cost of default, nearly1–2 %, and cost of operations, nearly 6–18 % of interest rates (Chakraborty 2010).

10Income pertaining to interest, fees and other service income from loans and investments.11Costs related to operations, including all personnel, depreciation and amortization, and adminis-trative costs.12Costs related to fund raising, including all interests and fees for any financial liability.13Provisions made to comply with some sort of regulation, either self-imposed or mandated byregulators on due loans, which are at the risk of default.14More details on the sustainability challenges faced by MFIs are discussed in literature reviewchapter.

Page 30: Sustainability of Indian Microfinance Institutions ||

1.5 An Overview of the Operational Self-Sustainability of Microfinance Institutions in India 13

0.00% 20.00% 40.00% 60.00% 80.00% 100.00% 120.00% 140.00% 160.00% 180.00% 200.00%

50 I

ndia

n M

icro

fina

nce

Inst

itut

ions

Operational Self-Sustainability Ratio for 2009

Fig. 1.3 Operational self-sustainability ratios of Indian microfinance institutions for the year 2009

The operational sustainability metrics of the vast majority MFIs remain unknownin India. NABARD substantiates this fact by portraying majority of Indian MFIs tobe opaque, remaining largely unproven in terms of its sustainability. This being thescenario, there is hardly been any empirical research done towards understandingthe sustainability of Indian MFIs. The seminal work that empirically analysed theoperations of 42 Indian MFIs for the year 2003 was undertaken by Crombruggheet al. (2008). In this work, the sample MFIs showed an average OSS ratio of 72 %,which is much below the cent per cent criteria. As this low ratio of 72 % wasreported by the relatively more institutionalized and sustainability conscious IndianMFIs, (that disclosed data to Sa-Dhan, an Indian MFI association), the OSS statusof the vast majority that did not disclose their data is presumed to be even lower.

To gain a closer view of the recent trends in the OSS status of Indian MFIs,available data is sourced from 50 Indian MFIs, that have disclosed their OSS metricsto Microfinance Information Exchange (MIX), for the period 2009. These MFIsseem to differ widely on their OSS metrics as on 1 April 2009, as shown below inFig. 1.3.

Page 31: Sustainability of Indian Microfinance Institutions ||

14 1 Introduction

The mean OSS of these 50 MFIs for the year 2009 is 117.06 %. This valuedepicts that in due course of time, most of the MFIs, have attained an operationallyself-sustainable status. But the microfinance crisis that occurred in India in theyear 2010 looks at this impressive OSS status with suspicious eyes. The spate ofsuicides among the microfinance clients in the Indian district of Andhra Pradesh,during the month of September 2010, allegedly due to exorbitant interest rates,levied by some of the MFIs, has resulted in a crisis in Indian Microfinance Industry,raising doubts on the sanctity of these impressive OSS rates. There have beenwidespread scepticisms as to whether Indian MFIs are managing their operationallyself-sustainable status by charging unreasonable interest rate from the clients.

1.6 Statement of Research Problem

To pre-empt Indian MFIs from augmenting their sustainability by charging unrea-sonable interest rates, a ceiling on interest rates was imposed on Indian MFIsin the year 2011. A special sub-committee appointed by Reserve Bank of India(RBI), called Malegam Committee (2011), fixed 26 % as the reasonable interestrate for Indian MFIs. Though this cap on interest rate is imposed, in literature it isobserved that an MFI can be said to levy reasonable interest rate, not just by thefact that it is charging low cost-covering interest rate. It should also ensure that itscosts are not excessively high due to inefficiencies that can be trimmed off fromits operating structure. Therefore, in this study we identify operationally efficientIndian MFIs which remained sustainable by charging the reasonable interest rateof 26 % or lower, from the clients, even prior to the happening of a crisis andceilings imposition. The intent is to understand how these operationally efficientIndian MFIs, which remained sustainable by charging a reasonable interest rate, aremanaging the determinant and discriminant factors of OSS. The strategies used bythese MFIs for managing these factors are documented for the reference of otherMFIs operating in the sector. A discussion on sustainability mismanagement issues,which can happen when MFIs become overly conscious about OSS, forgetting thelarger picture of client welfare, is also drawn with reference to Indian microfinancecrisis. This is done to ensure that the sector never loses its focus on social goals, inits pursuit for sustainability.

1.7 Research Aim and Objectives

In this book, the author intends to investigate about the issues related to sus-tainability of Indian MFIs, using a mixed-methods approach. For the purposeof this investigation, firstly, the author intends to identify the determinant anddiscriminant factors of the OSS of MFIs in India using quantitative analysis. Anon-parametric analysis is then undertaken to see how operationally efficient MFIs

Page 32: Sustainability of Indian Microfinance Institutions ||

1.8 Database, Sample and Methodology 15

in India which did remain sustainable by charging a reasonable interest rate fromthe poor clients, were managing these factors, even before the onset of the crisisand ceiling imposition. The strategies used by these MFIs to manage the factorsaffecting and discriminating OSS are documented for the reference of relatively lessefficient Indian MFIs. This is expected to aid these MFIs to enhance the efficiencyand sustainability of their microfinance operations. Thereafter, a discussion on thedangerous involved in MFIs mismanaging the sustainability factors is brought forth,taking Indian microfinance crisis as a reference. This is expected to serve as a gentlereminder that, being overly conscious of sustainability, oblivious of client welfarewould only amount to working against the spiritual foundation of the microfinancesector.

The research objectives pursued in this study are enumerated below in the orderin which they are undertaken:

(a) Identification of factors affecting the OSS of Indian MFIs: To identify thefactors that affect the OSS of Indian MFIs.

(b) Identification of factors discriminating the OSS status of Indian MFIs: Toidentify the factors that can discriminate and predict the OSS status of IndianMFIs.

(c) Identification of the efficient and sustainable Indian MFIs: To arrive at a setof efficient and sustainable Indian MFIs that can be benchmarked as peersor reference groups for other MFIs operating in Indian microfinance industry.These are the set of efficient Indian MFIs, which remain sustainable by charginga reasonable interest rate from the clients.

(d) Management of the factors affecting and discriminating the OSS status of IndianMFIs: To understand how the identified efficient and sustainable Indian MFIs,are managing the factors affecting and discriminating their OSS status and toseek confirmation from these MFIs on the relationship shared by these factorswith OSS. Also to reflect on the dangers involved in the mismanagement ofthese factors, by taking Indian microfinance crisis as reference.

1.8 Database, Sample and Methodology

As discussed in the earlier section, since most of Indian MFIs are opaque withrespect to their operational data, there exist data constraints for conducting microfi-nance research on large sample size. But from 2005 onwards the situation slightlyimproved and most of the institutionalized Indian MFIs began to share their data toMIX, a database maintained for MFIs by a nonprofit organization headquartered inWashington, DC (http://www.themix.org/).15

15The data from MIX is widely used in microfinance research. In terms of reliability of thedata, the organization’s claim is as follows: “MIX Market has a deep historical dataset, trackingindustry development since the 1990s. MIX sources data from audits, internal financial statements,

Page 33: Sustainability of Indian Microfinance Institutions ||

16 1 Introduction

A sample size of 50 MFIs was found to be the maximum number of IndianMFIs which has disclosed their operational data to MIX for the period 2005–2009. Therefore, for the purpose of this study, the sample comprises of these 50MFIs.

The methodology used in this work is a mixture of both quantitative andqualitative research methods. It involves a quantitative analysis to identify thefactors affecting and discriminating the OSS status of Indian MFIs and a qualitativeanalysis to understand how efficient and sustainable MFIs are managing thesefactors. For the initial quantitative analysis, (using multiple regression analysis andmultiple discriminant analysis), secondary data is sourced on 50 Indian MFIs fromMIX database over the period 2005–2009. An intermediary phase is then usedto link the quantitative analysis with the subsequent qualitative analysis. A non-parametric data envelopment analysis technique is conducted in this intermediaryphase, on the same sample of 50 Indian MFIs to identify the participant MFIs (i.e.efficient and sustainable peer MFIs) for the subsequent qualitative analysis. For theensuing qualitative analysis, primary data is collected by interviewing four efficientand sustainable Indian MFIs. Finally, the discussion is closed by citing literature onsustainability mismanagement issues, taking Indian microfinance crisis as reference.

As the sequence of this research is in the lines with an explanatory researchwork, a sequential explanatory mixed methods research design, that involves bothquantitative and qualitative analysis is framed to attain the research objectives. Moredetails on this research design are given in Chap. 3.

1.9 Scope of the Research Work

The focus of this research is to understand the issues related to the sustainabilityof Indian MFIs. With this aim the study firstly identifies the factors affecting anddiscriminating the OSS status of Indian MFIs, through a quantitative analysis.Thereafter, by adopting a non-parametric analysis technique, it identifies efficientMFIs, which remain sustainable by charging a reasonable interest rate from theclients. Finally, by pursuing a qualitative inquiry, it understands how these identifiedefficient and sustainable MFIs are managing the factors affecting and discriminatingthe OSS status of Indian MFIs. Thus, the study documents the managerial strategies

management reports or other documents and complements this data with questions directly to theMFI. MIX analysts and partners enter all data into the database; all data is reviewed by MIX staffand validated against a set of business rules before publication. Users can view and downloadsource documents directly from MIX Market to perform their own validation of the data enteredby MIX. Interim data in most cases is unaudited, by definition, but MIX cleans its data extensivelyusing a data audit system with over 150 audit rules that help analysts focus on the right issuesand follow-up with MFIs when necessary. Audit rules cover factors such as whether financialstatements balance or whether ratios levels are abnormally high or low for an MFI”. [Source:http://www.mixmarket.org/about/faqs/]

Page 34: Sustainability of Indian Microfinance Institutions ||

1.10 Organization of the Book 17

used by these efficient and sustainable MFIs, with respect to these significantdeterminants and discriminants of their OSS status. Since these strategies are usedby the efficient MFIs which operate sustainably at a reasonable interest rate, it canserve as a reference to other MFIs operating in the industry. The discussions inthis study are thus limited to the managerial aspects of the significant determinantand discriminant factors identified in the quantitative phase of the study. Moreoverthe efficient and sustainable peer MFIs identified in this study are arrived on thebasis of a relative efficiency and sustainability assessment undertaken on a sampleof 50 Indian MFIs, using parameters that are internal to the MFI’s operations.Operationally efficient MFIs which remain sustainable by charging an interest rateof 26 % or lower and which are comparators to other MFIs in the industry areregarded as efficient and sustainable peers. The rates the MFIs publish with MIXand in their annual reports, which are inclusive of interest and all fee expenses, istaken as the interest rates levied. No further adjustment for hidden costs allegedby media is made, as accounting for assumed costs would make standardizedcomparisons across MFIs difficult. Therefore, the focus is only on how the efficientand sustainable MFIs are managing their OSS, given the interest rate that theylevy from the clients. Thereafter a literature-based discussion is undertaken usingmicrofinance crisis as reference to see how mismanagement of OSS can be a threatto client welfare.

1.10 Organization of the Book

This book is divided into eight chapters. In the first chapter, an overview of theconcept of microfinance is presented. The research problem and research objectivespursued in this study are discussed in this introductory chapter. The remainingchapters are arranged as follows: Chap. 2 presents the literature review undertakento derive the research problem and research objectives. Chapter 3 explains thethree-phased sequential explanatory mixed-methods research design formulated inthis study, to fulfil the research objectives. Chapter 4 presents a discussion on thequantitative phase of this study, which fulfils the first and second research objectives.Chapter 5 presents a discussion on the intermediate participant selection phase,which fulfils the third objective of the study. Chapter 6 presents a discussion onthe qualitative phase, which investigates the managerial issues of MFIs, therebyfulfilling one aspect of the fourth objective of the study. Chapter 7 is the penultimatechapter where the author reflects on the dangers in mismanaging sustainabilityfactors, taking Indian microfinance crisis as a reference, to complete the fourthobjective. With this the qualitative phase of the study gets completed. Finally, Chap.8 concludes this study by presenting a summary of the findings drawn from each ofthe three analytical phases pursued in this study. The implications of these findingsare discussed as the key contributions made by this study. The limitations and scopefor future work are also discussed.

Page 35: Sustainability of Indian Microfinance Institutions ||

18 1 Introduction

References

Agarwal A (2008) The need for financial inclusion with an Indian perspective. Industrialdevelopment bank of Indian document. Economic Research IDBI, Mumbai

Armendariz de Aghion B, Morduch J (2005) The economics of microfinance. MIT Press,Cambridge

Bakhtiari S (2006) Microfinance and poverty reduction: some international evidence. Int Bus EconRes J 5(12):65–70

Basu P (2008) A financial system for India’s poor. In: Karmakar KG (ed) Microfinance in India.Sage, New Delhi, pp 19–32

Brau CJ, Woller GM (2004) Microfinance: a comprehensive review of existing literature. JEntrepren Financ Bus Ventur 9(1):1–26

Chakraborty M (2010) Room available to cut interest rates. The Economic TimesChristen RP, Rhyne E, Vogel R, McKean C (1995) Maximizing the outreach of microenterprise

finance: an analysis of successful microfinance programs. U.S. Agency for InternationalDevelopment (USAID) Program and Operations Assessment Report No. 10. Washington, DC

Consultative Group to Assist the Poor (CGAP) (2006) The new vision of microfinance: financialservices for the poor. www.networkers.org/.../The%20History%20of%20Microfinance.doc.Retrieved 25 Dec 2011

Crombrugghe A, Tenikue M, Sureda J (2008) Performance analysis for a sample of microfinanceinstitutions in India. Ann Pub Coop Econ 79(2):269–299

Ghatak M (2000) Screening by the company you keep: joint liability lending and the peer selectioneffect. Econ J 110:601–631

Ghatak M, Guinnane TW (1999) The economics of lending with joint liability: theory and practice.J Dev Econ 60:195–228

Gibbons DS, Meehan JW (2002) Financing microfinance for poverty reduction. http://www.microcreditsummit.org/papers/financing.pdf. Retrieved 18 Dec 2010

Intellecap (2010) Indian microfinance crisis of 2010: turf war or a battle of intentions? AnIntellecap white paper. Intellecap, Hyderabad

Johnson S, Rogaly B (1997) Microfinance and poverty reduction. Oxfam/Action Aid,Oxford/London

Ledgerwood J (1999) Microfinance handbook: an institutional and financial perspective. WorldBank, Washington, DC

Leeladhar V (2007) Indian financial sector reforms. Annual Washington conference of the Instituteof International Bankers. Text of speech. Institute of International Bankers, Washington, DC

Malegam Committee Report (2011) Report of the Reserve Bank of India sub-committee of itsCentral Board of Directors to study issues and concerns in the Micro Finance Institutions(MFI) sector. Reserve Bank of India. http://www.rbi.org.in/scripts/BS_PressReleaseDisplay.aspx?prid=23780. Retrieved 25 Feb 2010

Morduch J (1999) The microfinance promise. J Econ Lit 37(4):1569–1614Mutua K, Nataradol P, Otero M, Chung B (1996) The view from the field: perspectives from

managers of microfinance institutions. J Int Dev 8:179–193Narasimham M (1991) Report of the committee on the financial system. Reserve Bank of India,

MumbaiNational Council for Applied Economic Research (NCAER) (2011) Assessing the effectiveness of

small borrowings in India. NCAER–Centre for Macro Consumer Research, New DelhiOikoCredit (2005) Small loans great change: building a future with microfinance. In: International

microfinance symposium, BonnOtero M (2000) Bringing development back to micro finance. J Microfinanc 1(1):8–19Raj KN (1974) Monetary management and nationalization of banking in India. In: Mitra A (ed)

Economic theory and planning: essays in honour of A K Dasgupta. Oxford University, CalcuttaReserve Bank of India (RBI) (1954) Report on the all India rural credit survey of 1951–52. RBI,

Mumbai

Page 36: Sustainability of Indian Microfinance Institutions ||

References 19

Reserve Bank of India (RBI) (1995) Expert committee on integrated rural development programme(IRDP). RBI, Mumbai

Rhyne E (1998) The Yin and Yang of microfinance: reaching the poor and sustainability.MicroBank Bull 2(1):6–8

Robinson SM (1995) The paradigm shift in microfinance: a perspective from HIID. Paper presentedat the Harvard Institute for international development history conference. Harvard Institute forInternational Development, Bermuda

Robinson SM (2001) The microfinance revolution: sustainable finance for the poor. World Bank,Washington, DC

Ruthven O (2001) Money mosaics: financial choice and strategy in a West Delhi Squattersettlement. Finance and Development Research Programme Working Paper Series 32, IDPMUniversity of Manchester. http://www.man.ac.uk/idpm/. Retrieved 20 Feb 2009

Schreiner M (1996) Thinking about the performance and sustainability of microfinance organiza-tions. http://citeseerx.ist.psu.edu/viewdoc/summary. Retrieved 9 May 2010

Schreiner M, Colombet HH (2001) From urban to rural: lessons for microfinance from Argentina.Dev Policy Rev 19(3):339–354

Sen A (1999) Development as freedom. Anchor Books, New YorkSimanowitz A (2002) Ensuring impact: reaching the poorest while building financially self-

sufficient institutions, and showing improvement in the lives of the poorest women and theirfamilies. In: Daley-Harris S (ed) Pathways out of poverty: innovations in microfinance for thepoorest families. Kumarian Press, Bloomfield

Sinclair S (2001) Financial exclusion: an introductory survey. CRSIS/Heriot-Watt Univer-sity, Edinburgh. http://www.crsis.hw.ac.uk/Financialpercent20Exclusionpercent20Review.pdf.Retrieved 1 Sept 2009

Sinha S, Patole M (2003) Microfinance and the poverty of financial services: a perspective fromIndian experience. S Asia Econ J. http://sae.sagepub.com/content/4/2/301. Retrieved 27 Jul2010

Sriram MS, Upadhyayula SR (2004) The transformation of micro finance sector in India:experiences, options & future. J Microfinanc 6(4):89–112

Srnec K (2007) Microfinance—tool for poverty alleviation in the less developed countries.Habilitation Thesis. Institute of Tropics and Subtropics, Czech University of Life Sciences,Prague

Thorat YSP (2006) Microfinance in India: sectoral issues and challenges. Towards a sustainablemicrofinance outreach in India. NABARD, GTZ and SDC, New Delhi, pp 27–42

Thorat U (2007) Financial inclusion—the Indian experience. Financial inclusion conference, textof speech. DFID, London

UNCDF (United Nations Capital Development Fund) (2006) Building inclusive financial sectorsfor development. United Nations, New York

UNIFEM (United Nations Funds for Women) (2001) Micro-finance for women’s empowerment: aparticipatory learning, management and action approach. United Nations, Milton Keynes

Vaidyanathan A (2004) Task force on revival of cooperative credit institutions. Reserve Bank ofIndia, Mumbai

Yaron J, Benjamin MP, Piprek GL (1997) Rural finance: issues, designs and best practices,Environmentally and socially sustainable development studies and monographs series. WorldBank, Washington, DC

Page 37: Sustainability of Indian Microfinance Institutions ||

Chapter 2Literature Review: The Sustainability Debate

2.1 Preface

The previous chapter introduced the microfinance concept in general and focusedon the OSS of Indian MFIs in particular. It portrayed a general overview of theresearch investigation and presented the research objectives pursued in this work.Against this backdrop, this chapter presents a detailed review of the existing debatein microfinance literature regarding OSS of MFIs. With this literature review asa preface, in this chapter the author discusses how this study is expected to addvalue to existing microfinance literature. This will serve as a prelude to the researchinvestigation pursued.

2.2 Literature Review

Past research works that have studied the sustainability aspects of MFIs and thatare relevant to the research investigation are categorized in this chapter under thefollowing Sects. 2.2.1, 2.2.2, 2.2.3 and 2.2.4. The purpose of undertaking thisliterature survey is to narrow down to the research gaps addressed in this study.After narrowing down to the research gaps, the research objectives are revisited inthis chapter, by discussing the literature associated with them. Detailed discussionson literature specific to these objectives are presented in the respective chapters, inwhich each of these objectives is pursued.

2.2.1 Sustainability Definition and Metrics

According to Thapa (2007) sustainability in microfinance could relate toorganizational, managerial or financial sustainability. Mahajan and Nagasri (1999)

N. Marakkath, Sustainability of Indian Microfinance Institutions:A Mixed Methods Approach, India Studies in Business and Economics,DOI 10.1007/978-81-322-1629-2__2, © Springer India 2014

21

Page 38: Sustainability of Indian Microfinance Institutions ||

22 2 Literature Review: The Sustainability Debate

opine that microfinance sustainability can be related to mission sustainability,financial sustainability, demand sustainability, legal and regulatory framework andownership and governance of MFIs. Among the different forms of sustainability,the one that has received the most attention in microfinance literature is financialsustainability. Financial sustainability is denoted in microfinance literature bythree prominent metrics—operational self-sustainability ratio (OSS), financial self-sustainability ratio (FSS) and Subsidy Dependence Index (SDI).1 Of these the mostbasic measure of financial sustainability of an MFI is OSS. OSS is a subsidy-independent accounting measure of sustainability which denotes the ability of MFIto earn revenue to cover its costs and reach the poor now and in future (Schreiner1996). More specifically, it is the ability of MFI to generate enough income fromits operations to cover its financing costs, operating costs and loan loss provisions,regardless of whether it is subsidized or not (Meyer 2002). Such an operationallysustainable MFI is eventually expected to attain an FSS status, a higher state thatdenotes the ability of an MFI to generate enough self-generated revenue to coverits financing costs, operating costs and cost of provisions for losses, without resortto subsidies (Meyer 2002; Sharma and Nepal 1997; Gibbons and Meehan 2002;Ledgerwood 1999; Rosenberg 2009). According to Pissarides et al. (2004), FSSstatus for an MFI is a coveted status in which an MFI operates, without any usageof subsidies, grants or other concession resources to profitably provide financeto the poor. Akin to FSS, SDI is another subsidy-dependent sustainability metricfor MFIs, developed by Yaron (1992). SDI is a ratio that indicates the percentageincrease required in on-lending interest rates to completely eliminate all subsidiesreceived by an MFI. Calculation of both SDI and FSS requires the computation ofthe amount of concessional funds and its associated opportunity cost for the MFI.Literature therefore regards OSS to be a more reliable measure of sustainabilitythan SDI or FSS, given the difficulty in accurately estimating these two figures(Crombrugghe et al. 2008). More details on the operational differences betweenthese metrics are given in Appendix 1.

Apart from this empirical accuracy, OSS is a financial sustainability mea-sure which can assess the sustainability of MFIs, without discriminating forthe usage of subsidies (Nyamsogoro 2010). MFIs belonging to either school ofthought—‘welfarist MFIs belonging to Poverty Lending School of Microfinance’and ‘Institutionalist MFIs belonging to Financial System School of Microfinance’—can be compared using OSS ratio, because it is a subsidy-independent sustainabilitymeasure. Welfarist MFIs are those MFIs that belong to the school that believes indepth of outreach. These MFIs believe in reaching out to the very poor clientele,with or without subsidy assistance (Bhatt and Tang 2001). They reach out to

1Return on Asset (ROA) is also used a metric for sustainability assessment. But since literaturedoes not clearly demarcate the level of ROA, at which an MFI can be assumed to be sustainable, itis used less in relative sustainability assessments.

Page 39: Sustainability of Indian Microfinance Institutions ||

2.2 Literature Review 23

the extremely poor clientele by operating in the harsh socio-economic conditionsand geographically isolated communities, where the very poor actually dwell.Therefore, they are justified in resorting to subsidies initially as catalysts, tonurture its operations (Woller et al. 1999; Annapurna 2008). These MFIs will takeconsiderably lengthier time period to reach the coveted status of FSS or SDI, thanthe institutionalist MFIs. Institutionalist MFIs are those that belong to the schoolthat believes in breadth of outreach and financial self-sustainability. These MFIsbelieve in reaching out to maximize number of economically weak clientele, ina self-sustainable manner, without the use of any subsidy support (Woller andWoodworth 2001; Olivares-Polanco 2005). Since the MFIs belonging to the twodifferent schools of welfarism and institutionalism cannot be treated as comparablesusing FSS and SDI ratio, the OSS ratio is preferred as a more reliable measure forrelative sustainability assessment in this study. FSS and SDI, though a desirablestatus of sustainability for MFIs, are often a difficult proposition for welfarist MFIs,who strictly hold on the spiritual foundation of microfinance. Since we value thepoverty-lending philosophy followed by welfarist MFIs, it is decided to use asustainability metric that does not have any bias for MFIs based on the usage ofsubsidies. As Woller et al. (1999) puts it, what matters is how subsidies are usedand not whether subsidies are used or not. Therefore, the sustainability metrics thatdiscriminates MFIs on the basis of usage of subsidies is not used in this study.Instead the basic metric of OSS is used. The challenges faced by MFIs to attainOSS are surveyed in next section.

2.2.2 Challenges Faced by Microfinance Institutionsin Attaining Sustainability

MFIs operate in an environment of high information asymmetric credit marketrisk (Ross and Savanti 2005). Information asymmetric risk arises in credit-lendingtransactions, as the lender has less information about the creditworthiness of theborrower than the borrower himself. Such risks are all the more exacerbatedin microfinance market as the poor borrowers lack credit history. Informationasymmetric credit market risks denotes the ex ante risk of adverse selection,2 interim

2Adverse selection risk arises when the lender has poor information about the borrowers whilenegotiating the credit-lending transaction. With the limited information on the poor borrowers, thelender cannot screen the riskier borrowers from safer ones. Therefore, there is an adverse selectionrisk of lending to the more risky borrowers.

Page 40: Sustainability of Indian Microfinance Institutions ||

24 2 Literature Review: The Sustainability Debate

risk of moral hazard3 and the two ex post risks of costly audits and enforcement4

(Akerlof 1970; Scholtens and Wensveen 2003; Stiglitz and Weiss 1981).MFIs mitigate these information asymmetric credit market risks—adverse selec-

tion, by affecting group formation among the poor borrowers with joint liability;moral hazard, by inducing group members to influence the way other membersselect their projects; costly monitoring, by helping the lender avoid external audits;and enforcement problems, by encouraging borrowers to repay their loans withoutthe lender having to impose sanctions—by its unconventional group-lending models(Ghatak and Guinnane 1999; Ghatak 2000).

But the group-lending model used by MFIs to mitigate these risk results inhigh intermediation costs (i.e. operating costs) for the MFIs (Thorat 2006; Savita2007). The group-lending models entails peculiar costs, such as group formationcosts, costs of training the borrowers on the procedures, cost of higher degree ofsupervision and higher frequency of instalment payments, all adding to the operatingcosts of the MFI. Moreover since the average microfinance loan size is small, thetransaction cost on a percentage basis for such microfinance loan tends to be higher.Thus, the high operating costs incurred by MFIs are a major challenge at the stakeof its sustainability. To cover these high costs by generating a surplus and remainoperationally self-sustainable is a formidable task for MFIs. Though a difficult task,attaining sustainability is considered to be imperative for an MFI. Notable worksthat examined the relevance of the sustainability for MFIs are discussed in nextsection.

2.2.3 Relevance of Sustainability for Microfinance Institutions

Rhyne (1998) opines that sustainability is the means to the goal of outreachto the poor. Rhyne cites that only by achieving sustainability will microfinanceprogrammes gain access to the funding they need over time, to serve the hithertounreached poor clients.

Otero (2000) avowedly states that only if an MFI achieves this sustainability, willit have the potential to:

(a) Work towards the objective of poverty alleviation by rendering financial servicesto the poor

3Moral hazard risk arises because the lender has difficulty in monitoring the behaviour of thepoor borrowers once the loans are disbursed. Therefore, the lender does not know whether theloan is being used optimally for the intended purpose for which it is sanctioned. The lender lacksinformation about the performance of the credit-lending transaction and the probability for theloans disbursed to be misused, which results in the risk of moral hazard.4Costly audit and enforcement risks arise because it becomes too costly for the lender to audit andenforce payments on the small loans disbursed to the poor, which lack collateral support.

Page 41: Sustainability of Indian Microfinance Institutions ||

2.2 Literature Review 25

(b) Exist as permanent private financial institutions, which acts as a distributionalchannel to convert economic growth to improved well-being among the poor

(c) Deepen a nation’s financial system, by serving as a financially transparent andregulated institution

Meyer (2002) notes that sustainability is imperative for MFIs, as the poor needto have access to financial service on long-term basis rather than just a one-timefinancial support. Navajas et al. (2000) also confirm this by stating that short-termfinancial assistance would worsen the welfare of the poor. Burkett (2007) discussesthe primacy of sustainability of MFIs, by explaining the relationship betweenmicrofinance and neo-liberalism. According to the neo-liberal and neo-conservativeeconomic agenda, microfinance is a market-oriented solution to poverty alleviation.As per this ideology, microfinance, being a market entity, is required to eventuallywean themselves off from all forms of financial control imposed by the state, donorsor subsidies. Therefore, MFIs are structured to operate as social businesses that arepermitted to charge cost-covering interest rates and earn profits from self-generatedrevenue, in order to attract funds for their sustenance. Such profit orientation,sustainability consciousness and competition in the microfinance industry wereencouraged, so as to inculcate the spirit of cost minimization among the MFIs,leading to the provision of financial services to the poor at a reasonable interest rate.

Thus, the review of literature on the relevance of sustainability for MFIs depictsthat sustainability, though a nonprofit equivalent of profitability in microfinance, ispursued by MFIs not for the sake of profitability, but as a means to fulfil its socialobjective of poverty lending (Woller and Schreiner 2002; Rhyne 1998). This makesit clear that any efforts made by MFIs to attain sustainability, oblivious of this socialobjective of outreach to the poor, are meaningless.

2.2.4 Balancing the Dual Goals of Microfinance Institutions

Researchers hold disputing views, about the ability of an MFI to pursue the dualgoals of sustainability (financial goal) and outreach to the poor (social goal).According to (Christen et al. 1995) and Otero and Rhyne (1994), the two goals—outreach and financial sustainability of an MFI—are complementary to each other.This is so, as the number of clients increases, the MFI will experience economiesof scale. This in turn will result in cost-efficiency for the MFI, which will helpthem to attain financial sustainability. On the other hand, Hulme and Mosely (1996)argue that there is inverse relationship between outreach and financial sustainability.The rationale is that higher outreach for an MFI means more transaction cost inmitigating the information asymmetry among its clients, making it difficult forit to attain MFI financial sustainability. Conning (1999) confirms this trade-offbetween the goals of outreach and sustainability, after analysing data for 72 MFIs.According to Conning, trade-offs exists due to the costly monitoring and controlsystems required for the highly information asymmetric borrowers of MFIs, which

Page 42: Sustainability of Indian Microfinance Institutions ||

26 2 Literature Review: The Sustainability Debate

serve as substitutes for their collaterals. The challenges in pursuing the dual goalsof outreach and sustainability are also surveyed by Von Pischke (1996), Robinson(2001) and Armendariz de Aghion and Morduch (2005). Morduch (2000) discussesthe eclectic views that are associated with the trade-offs between these two goals ofsustainability and poverty alleviation through outreach, as microfinance schism. Amore recent study by Cull et al. (2007) empirically confirms the trade-off aspectby examining the outreach and profitability data for 124 institutions across 49countries.

Thus, the above literature-based discussion shows that there is no conclusiveevidence to prove the proposition that financial sustainability and poverty alleviationthrough outreach are complementary to each other. According to Simanowitz(2002) while the microfinance industry has established guidelines for measuringand reporting financial sustainability, there are no established similar standards forassessing the impact of outreach in its true sense. The prevalent methods, used bymicrofinance researchers to assess the impact of the microfinance programme ontheir clients, are, by measuring changes in dependent variables such as the levelof income, the level of production, sales, assets or the general well-being of theclients, ex ante and ex post microfinance interventions (Alfaro 1999; Bhatt andTang 2001). The underlying assumption in these methods is the existence of adirect causal relationship between the credit disbursed through microfinance and theobserved change in the dependent variables listed above (Rhyne 1994). The maincriticisms of this methodology is that the results from such studies face problemsof selection bias, lack of control groups and paucity of longitudinal data (Bhattand Tang 2001). Olivares-Polanco (2005) points out additional flaws in such impactassessments. Since money is fungible in nature, when a direct casual relationshipis assumed between credit and other dependent variables, proper controlling forrest of the sources and uses of funds and other factors that have an effect ondependent variable must be made. Olivares-Polanco observes that even thoughmost impact analysis includes possible control groups for some of these variables,the problem of equivalence between the control group and experimental group(the group that actually receiving the loans) is an issue that affects the accuracyof most impact assessments. This being the case, even though an MFI has animpressive financial sustainability ratio, researchers agree that nothing conclusivecan be said about its ability to alleviate poverty through outreach. This is so,as there exist data and methodological limitations in capturing the effect of afungible commodity like credit on the poor. Owing to these reasons, it began tobe widely assumed in the industry that ‘more microfinance’ can be substitutedfor ‘more poverty reduction’ (Bateman and Chang 2009). Gradually the successof MFIs began to be judged widely by their ability to be financially sustainable,and the poverty reduction objective was assumed to be achieved concomitantlywhen microfinance services are made available to the poor. Several studies wereundertaken to understand the determinants of sustainability, which would enableMFIs to focus on their sustainability. Notable studies in this direction were byAdongo and Stork (2005), Hartarska and Nadolnyak (2007), Crabb (2008), Ayayiand Sene (2007) and Crombrugghe et al. (2008). Of these studies the one undertaken

Page 43: Sustainability of Indian Microfinance Institutions ||

2.2 Literature Review 27

by Crombrugghe et al. (2008) in Indian context tested whether the pursuit ofsustainability would result in mission drift for the MFIs. Though the study concludesby stating no evidence for such a mission drift, the results recommend the practice oftargeting interest rates levied from customers of MFI to augment MFI sustainability.The emphasis on sustainability thus made the practice of charging cost-coveringinterest rates widely acceptable for the microfinance industry. With the tenet ofsustainability guiding the sector, the MFIs also began to expand its scale. Butthe proposition that ‘more microfinance would result in more poverty alleviation’became widely sceptical due to the commercialized trends that became prevalentamong MFIs. Christen and Drake (2002) regard commercialized microfinance tobe characterized by an MFI’s usage of market-based principles in its operations.Commercialized microfinance with the pressure to meet the return expectations ofdebt and equity investors began to influence the behaviour of MFI loan officers.Loan officers of the commercialized MFIs began to actively seek new clients andprefer frequent repayment schedules and high interest rates and repayment rates.Compartamos, a commercialized Latin American MFI that was first to go for anInitial Public Offering in the year 2007, was reported to levy an interest rate closeto 100 % per annum. The ethicality of this practice has been widely deliberated byacademicians and practitioners (Karnani 2011; Ashta 2011; Lewis 2008; Rhyne andGuimon 2007; Rosenberg 2007).

Sinha (2010) observes such commercialized practices to have led to bullish trendsamong global MFIs, marking a growth rate of 70–100 % per annum. This mademicrofinance inevitably a market that is flooded by profit seekers. This resulted invulnerabilities in a number of microfinance markets, especially in Bosnia, Morocco,Nicaragua, Pakistan and India. In India, the world’s largest microfinance market, acrisis was triggered by a combination of the highly successful stock market listingof India’s largest MFI, Swayam Krishi Sangam (SKS), which projected the extendof profits made by microfinance businesses, alongside several cases of suicidesamongst the clients of MFIs in the district of Andhra Pradesh (Panwar 2011). Thespate of suicides was allegedly due to exorbitant interest rates and the coerciverecovery practices adopted by some firms masquerading as MFIs in Andhra Pradesh.Though there is no systematic investigation and conclusive evidence for the suicidesto be instigated by MFI activities, these episodes during the month of September2010 threatened the viability of the entire microfinance sector in India (Intellecap2010; Swami et al. 2010). As a result Indian MFIs, which are reputed globally asthe least cost players in the microfinance industry, began to face reputation risks.Their operations were attributed to be tantamount to that of moneylenders, whocharge exorbitant interest rates and use coercive recovery practices to exploit thepoor (Sinha 2010). The Malegam Committee, a special sub-committee appointedby RBI during the post-crisis period, to study the issues and challenges of thesector, cited few large Indian MFIs to be levying interest rates close to 50.53 %.The committee’s report submitted in 2011 cited that on an average the interest ratecharged by Indian MFIs came to 28–36 % in the year 2009–2010, providing themwith huge financial margins. As per the report, it provided a financial margin closeto 24 % for large Indian MFIs, whose average financing cost is 11.78 % and interest

Page 44: Sustainability of Indian Microfinance Institutions ||

28 2 Literature Review: The Sustainability Debate

rate yield is 36.79 % for the year 2009–2010. Even the small Indian MFIs reaped afinancial margin of 16 % as their average financing cost was 11.71 % and interestrate yield was 28.73 % for the year 2009–2010. Though these interest rates wereless compared to the global average of 31 %, for a nation like India, this pricing wasviewed as more commercialized than being pro-poor (Sinha 2010). Ashta (2011)cites that even the global average MFI interest rate of 28 %, as reported by CGAP forthe year 2006, would seem unethical in the eye of donors from developed countrieswho are witnessing near zero interest rates.

In addition to the deliberations on the apparent interest rates levied by MFIs,Rosenberg et al. (2010) projects yet another subtle lens for viewing the ethicalside of MFI’s pricing. Their study conducted on 555 MFIs across the world, toinvestigate whether the poor are being exploited by the high microcredit interestrates charged by the MFIs, reported that though apparently MFI interest rates seemhigher than that charged by commercial banks, there is no evidence to prove theseinterest rates to be exploitative and abusive in nature. But the study admits that asthe extent of inefficiencies that could be trimmed off from an MFI’s operating costis not assessed, nothing conclusive can be commented on the interest rates chargedby the MFIs. This is so because interest rates are to be considered unreasonable,if it is loaded with avoidable inefficiencies in an MFI’s operating structure. Theauthors add that though the practice of charging cost-covering interest rate is widelyadvocated in microfinance industry, levying of unreasonable interest rate to augmentsustainability is not an acceptable practice in this industry.

But what constitute a reasonable interest rate have always been a complex issueand a debatable matter in the industry. Rosenberg et al. (2010) observe that ‘Even aninterest rate that only covers costs and includes no profit can still be unreasonableif the costs are excessively high because of avoidable inefficiencies’. Ashta (2009)opines that the interest rate should cover the minimum required by the MFI to stayin business and must be bound by the maximum affordable to clients to add welfare.In his view, ‘The question of ethics is how the interest should be shared along thisspectrum. This sharing of the margin needs to be assessed by social performancerating firms’. Yunus (2009) opines that ‘If you are being true to microfinance thenthere should not be more than 15 % difference in the interest rate at which youborrow and that at which you lend’. In India, as an aftermath of the crisis, toensure that the MFIs levy only a reasonable interest rate from the clients, the RBIimposed an interest rate and financial margin ceiling for MFIs at 26 % and 12 %,respectively. Karnani (2011) observes that such ceiling and regulation is imperativefor the industry, because the microfinance is a non-competitive market, which workswith financially illiterate and irrational clients. In such a market condition, withmonopoly rents and vulnerable customers, expecting competition to discover afair and reasonable price would only justify client exploitation. Therefore, Karnanifavours ceiling of interest rates in the sector. In his view the dangers of interestrate ceiling, in form of restricted financial access to poor, as foreseen by Fernando(2006) and Helms and Reille (2004), would not apply to a non-competitive andmonopolistic microfinance industry. But the perceptions of Indian MFI practitionerson the issue of interest rate caps were different. There were huge hues and cries

Page 45: Sustainability of Indian Microfinance Institutions ||

2.3 Summary of Inferences and Research Gaps Existing in Microfinance Literature 29

among Indian practitioners, with concerns as to whether small MFIs would beable to operate and remain sustainable at this rate (Samarapally and Gaul 2011).Deliberations on the correctness of having a regulatory cap to ensure reasonablepricing have resulted in mixed conclusions. Ashta and Assadi (2010) sums this byobserving that though free market lobbyists argue for removal of any caps, thereis no conclusive evidence to believe that absence of caps would foster growth inmicrofinance.

Thus, the above review shows that in India, owing to the crisis, the scenario issuch that MFI needs to learn how to operate sustainably by charging a fixed interestrate of 26 % or lower. The next section presents the research gap identified based onthis inference from literature.

2.3 Summary of Inferences and Research Gaps Existingin Microfinance Literature

Against the backdrop of the literature review undertaken in this chapter, it is inferredthat attaining sustainability is essential for an MFI to perpetually operate in thesector. But as depicted in literature, remaining sustainable by charging unreasonableinterest rates from the poor is not an acceptable practice in this industry. Incidenceof such practices has tarnished the social image of MFIs and has attracted regulatoryattention in the form of interest rate caps in Indian microfinance industry. Therefore,the way forward for Indian MFIs is to operate sustainably at the restricted interestrate ceiling or even lower. Only then will the industry be able to regain its lostreputation and be true to its role as a financial intermediary for the poor. In orderto do this, as Rosenberg et al. (2010) point out, it is not just enough that theMFIs charge a low cost-covering interest. It should also ensure that its costs arenot excessively high due to operational inefficiencies. The MFI managers shouldlearn to manage their MFIs sustainably by charging a price that reflects the benefitsof operational efficiency to the clients. This suggests that though Indian MFIs arerenowned to be low-cost players, there is still scope for the operationally inefficientplayers in the industry to trim off their wastage and reduce the pricing of their loans.By understanding their relative efficiency performance in the industry, Indian MFIscan determine how well they are utilizing their resources and where to minimizeinputs in their operations to improve their current performance. This aspect is barelyaddressed in literature. As reviewed in Sect. 2.2.4, empirical studies undertaken inthe direction of assessing the performance of MFIs have predominantly directedtheir focus on MFI’s sustainability. Contrary to the usual sustainability assessment,one study by Qayyum and Ahmad (2006) has addressed the efficient and sustainableperformance of MFIs. The results of the study identified six Indian MFIs as efficientand five out of them as efficient and sustainable in its operations. But the studyhas not assessed the dual goals of MFIs nor has it benchmarked the MFIs andassessed the interest rates levied by the efficient and sustainable MFIs. It has also not

Page 46: Sustainability of Indian Microfinance Institutions ||

30 2 Literature Review: The Sustainability Debate

documented the managerial practices used by them to attain sustainability. Moreoverthere has not been any discussion on how being overconscious of sustainability,oblivious of client welfare, can hamper the sector.

Therefore in this study, these research gaps are being addressed. The aim is tounderstand these issues related to the sustainability of Indian MFIs. As discussed inthe introductory chapter, four specific objectives are pursued to fulfil this researchaim. Using a mixture of quantitative and qualitative methods, this study fulfils theseobjectives. These objectives are revisited in the subsequent sections.

2.4 Research Objectives Revisited Along with AssociatedLiterature Review

In this section the four research objectives pursued in this study are revisited in thelight of the literature associated with it.

2.4.1 Objective 1: To Identify the Factors Affectingthe Operational Self-Sustainability of IndianMicrofinance Institutions

Morduch (1999) in his work, The Microfinance Promise, reiterates the need forempirical research in the field of microfinance that can aid in deriving the factorsaffecting the sustainability of MFIs. Several authors have worked in this direction.Adongo and Stork (2005) have attempted to identify the factors affecting the finan-cial sustainability of MFIs in Namibia. Factors, like donor dependency, group lend-ing and per capita income of group members, were found to have significant influ-ence on financial sustainability of MFIs. Hartarska and Nadolnyak (2007) found thesize of MFI to have a positive influence on the OSS of Bangladeshi MFIs and capitalratios to have a negative impact. Crabb (2008) found portfolio at risk, total numberof borrowers and economic freedom to have a significant influence on the OSS ofMFIs across the developing nations. Though this work identified relevant factors, itdid not discuss much about the managerial aspects of the significant factors.

Such an effort would have added more pragmatism to the issue under study.Ayayi and Sene (2007) filled this gap by testing the significance of a set ofmanagerial factors on the sustainability of 217 MFIs in 101 countries over theperiod of 1998–2006. The results of the regression analysis undertaken shows thata high-quality credit portfolio, coupled with the application of sufficiently highinterest rates that allow a reasonable profit and sound management, are instrumentalto the financial sustainability of MFIs. Though this study has taken care of themanagerial factors, it does not test the effect of average loan size per borroweron the sustainability of MFIs, a variable that proxies the depth of outreach as

Page 47: Sustainability of Indian Microfinance Institutions ||

2.4 Research Objectives Revisited Along with Associated Literature Review 31

per microfinance literature. Therefore, the study is silent on the trade-off effectbetween depth of outreach and sustainability as discussed by Von Pischke (1996).Crombrugghe et al. (2008) takes care of this aspect in a performance analysisconducted on 42 Indian MFIs. The results suggest that the challenge of coveringcosts on small and partly unsecured loans can be met, without necessarily increasingthe size of the loans. Better targeting of the interest rate policy, increasing thenumber of borrowers per field officer and usage of SHG model are recommended tomeet the cost on small loans and remain sustainable.

The above studies contribute to literature by identifying the significant factorsaffecting the sustainability of MFIs, in different contexts. But it does not assumea practical perspective and address how the practitioners should manage theseidentified factors, in order to enhance their MFI’s sustainability. Therefore, thestrategies that the best performers in the industry use to manage these factors arenot documented in literature. In this study, this research gap is addressed. Afteridentifying the factors affecting and discriminating the OSS status of Indian MFIs,the strategies used by efficient and sustainable MFIs to manage these factors will bedocumented for the reference of other MFIs in the industry.

2.4.2 Objective 2: To Identify the Factors Discriminatingthe Operational Self-Sustainability Status of IndianMicrofinance Institutions

No empirical studies have been made till date towards identifying the factors thatdiscriminate or predict the OSS status of Indian MFIs. This research gap will beaddressed in this study by pursing its second objective. As there are no past worksin microfinance in this direction, in order to identify the discriminants or predictorsof MFI sustainability, Altman’s (1968) celebrated model on bankruptcy predictionwas referred to.

2.4.3 Objective 3: To Identify Efficient and Sustainable IndianMicrofinance Institutions

Ledgerwood (1999) regards efficient performance of an MFI to indicate how wellan institution is managing its operations. Woller (2000) observes that attainingoperational efficiency is crucial for MFIs, as efficiency is an important factorcontributing to the self-sustainability of MFIs. CGAP (2003) adds to the needfor MFIs to achieve efficiency in operations, by stating that low efficiency inoperations can make interest rates levied by the MFI to be higher than necessary.Though achieving operational efficiency is pertinent for charging a reasonableinterest rate and for attaining sustainability, in literature there has been only one

Page 48: Sustainability of Indian Microfinance Institutions ||

32 2 Literature Review: The Sustainability Debate

study addressing this issue, in the Indian context.5 Qayyum and Ahmad (2006)has undertaken this study on MFIs based out in South Asia. In their study, 15Pakistani, 25 Indian and 45 Bangladeshi MFIs were assessed using a DEA model.This DEA model was tested only on a small sample of Indian MFIs, and it didnot capture the dual goals of an MFI. The efficiency analysis undertaken was notcomprehensive as it did not undertake a benchmarking analysis for the sample.Moreover it did not suggest the extent of input reductions that the lesser-efficientMFIs should achieve to enhance their efficiency. The subsequent sustainabilityassessment undertaken by Qayyum and Ahmad on the efficient MFIs captured onlythe growth dimension of sustainability. Therefore, an efficiency and sustainabilitymodel that is multidimensional, which is capable of benchmarking the MFIs andsuggesting the input minimization needed for the sample MFIs, have not beencarried out in Indian context. This research gap will be addressed in this study bypursing its third objective.

2.4.4 Objective 4: To Understand About the Managementof the Factors Affecting and Discriminatingthe Operational Self-Sustainability of IndianMicrofinance Institutions

According to Gonzalez-Vega (1998), the more efficient a firm is, the less is thegap between actual and best managerial practices in microfinance. Microfinancemanagerial best-practices literature suggest that in order to achieve sustainability,MFIs should reduce operating costs, drive up staff productivity, achieve significantscale and charge cost-covering interest rate (CGAP 1996; Woller 2000). Thoughthere has been widespread trend among MFIs to replicate practices associated withthese factors, across different countries, Woller and Schreiner (2002) observe thatall factors may not have same relevance across countries. Therefore, identificationof factors relevant to the sustainability of MFIs in specific countries is needed.Therefore in this study, an attempt is made to identify the factors that affect anddiscriminate the OSS of Indian MFIs. Thereafter, the strategies adopted by efficientand sustainable Indian MFIs to manage these factors are documented. Though inthis study we do not claim these strategies to be best practices, it is considered tobe worthy of reference and emulation for the other Indian MFIs to enhance theirefficiency and sustainability status. This is so because Microfinance Credit RatingInformation Ltd (M-CRIL), the global leader in the financial rating of MFIs and theleading credit rating agency for Indian MFIs, while analysing the MFI practicesof Indian MFIs in the year 2002, observes thus: ‘The more progressive MFIshave become successful partly through the development and adoption of practices

5Studies undertaken in other nations are reviewed in Chap. 5.

Page 49: Sustainability of Indian Microfinance Institutions ||

2.6 Summary 33

that are not only appropriate to their own particular conditions but also to thepractice of microfinance in general’. Having discussed these managerial strategies,it is also imperative to understand how mismanagement of the determinants anddiscriminants of OSS can go against the social goals of the MFIs. This is pertinentin a scenario where the sector is recovering from a crisis. So both management andmismanagement issues are discussed as fourth objective of the study.

2.5 Expected Value Additions from This Research Workto Microfinance Literature

This work is expected to add value to Indian microfinance literature by:

(a) Identifying the determinant and discriminant factors that Indian MFI managersshould concentrate on to enhance the OSS of their MFIs

(b) Formulating an OSS Predictor Model that predicts the OSS status of Indian MFI2 years from the date of estimation

(c) Formulating a Data Envelopment Analysis model that takes into account thedual goals of an MFI for assessing the efficiency of Indian MFIs

(d) Formulating a sustainability assessment model that assesses the four pertinentdimensions of an MFI’s sustainability

(e) Benchmarking the sample Indian MFIs to identify the set of efficient andsustainable Indian MFIs, which can serve as a peer or reference group for theother Indian MFIs

(f) Identifying the extent of input minimization to be achieved by Indian MFIs toenhance their operational efficiency

(g) Documenting the strategies used by selected efficient and sustainable MFIs inmanaging the factors affecting and discriminating their OSS status

(h) Presenting refections on the dangers involved in the mismanagement of OSS,using Indian microfinance crisis as reference

These contributions are expected to have practical implications for Indianmicrofinance managers as it will serve as a valuable learning experience for themto understand how to manage their MFI’s OSS. It is also expected to have policyimplications for Indian MFI regulators, in facilitating the management of OSS ofIndian MFIs without exploiting clients.

2.6 Summary

This chapter was dedicated to review the current debate in literature regardingan MFI’s pursuit of sustainability. The discussion in this chapter portrayed therelevance of pursing this research, against the larger canvas of microfinanceliterature. The expected value additions from this study were also enumerated.

Page 50: Sustainability of Indian Microfinance Institutions ||

34 2 Literature Review: The Sustainability Debate

With this literature support as a preface for this research investigation, the nextchapter explains the research design formulated for achieving the objectives of thisstudy.

References

Adongo J, Stork C (2005) Factors influencing the financial sustainability of selected microfinanceinstitutions in Namibia. Namibian Economic Policy Research Unit (NEPRU) Research ReportNo 38: 1026–9258

Akerlof GA (1970) The market for “Lemons”: quality uncertainty and the market mechanism.Quart J Econ 84(3):488–500

Alfaro GLN (1999) Sostenibilidady Alcance en Instituciones Financieras de Desarrollo Parala Micro y Pequeña Empresa: Un Modelo Aplicado a Tres Casos de Estudio (Ademi Inc.,BancoSol, S.A. y Génesis Empresarial). INCAE, Centro de Políticas Públicas: San José, CostaRica

Altman EI (1968) Financial ratios, discriminant analysis and the prediction of corporatebankruptcy. J Financ 23(4):189–209

Annapurna N (2008) Is the double bottom line attainable?—Context of interview with David SGibbons. IIMB Manage Rev 20(3):279

Armendariz de Aghion B, Morduch J (2005) The economics of microfinance. MIT Press,Cambridge

Ashta A (2009) Interest rate ethics: an aspect of social performance in microfinance. Paperpresented at European microfinance week, European microfinance platform, Luxembourg

Ashta A (ed) (2011) Advanced technologies for microfinance: solutions and challenges. IGIGlobal, Hershey

Ashta A, Assadi D (2010) Should online micro-lending be for profit or for philanthropy? DhanaXand Rang De. J Innovat Econ 2(6):123–146

Ayayi GA, Sene M (2007) What drives microfinance institution’s financial sustainability. J DevArea 44(1):303–324

Bateman M, Chang H (2009) The microfinance illusion. http://www.econ.cam.ac.uk/faculty/chang/pubs/Microfinance.pdf. Retrieved 21 May 2010

Bhatt N, Tang SY (2001) Delivering microfinance in developing countries: controversies and policyperspective. Policy Stud Organ 29(2):319–334

Burkket I (2007) Globalised microfinance economic empowerment or just debt? In: Dominelli L(ed) Revitalising communities in a globalizing world. Ashgate, Hampshire, pp 155–160

Christen RP, Drake D (2002) Commercialization. In: Drake D, Rhyne E (eds) The commercializa-tion of microfinance: balancing business and development. Kumarian Press, Bloomfield

Christen RP, Rhyne E, Vogel R, McKean C (1995) Maximizing the outreach of microenterprisefinance: an analysis of successful microfinance programs. U.S. Agency for InternationalDevelopment (USAID) program and operations assessment report no. 10. Washington, DC

Conning J (1999) Outreach, sustainability and leverage in monitored and peer monitoring lending.J Dev Econ 60:229–248

Consultative Group to Assist the Poor (CGAP) (1996) Microcredit interest rates. http://www.cgap.org/p/site/c/template.rc/1.9.2696/. Retrieved 25 Dec 2011

Consultative Group to Assist the Poor (CGAP) (2003) Making sense of microcredit interest rates.http://www.cgap.org/gm/document-1.9.2434/DonorBrief_06.pdf. Retrieved 25 Dec 2010

Crabb P (2008) Economic freedom and the success of microfinance institutions. J Dev Entrepren,World Scientific Publishing Co. Pvt. Ltd., 13(2):205–219.

Crombrugghe A, Tenikue M, Sureda J (2008) Performance analysis for a sample of microfinanceinstitutions in India. Ann Pub Coop Econ 79(2):269–299

Page 51: Sustainability of Indian Microfinance Institutions ||

References 35

Cull R, Demirguc-Kunt A, Morduch J (2007) Financial performance and outreach: a globalanalysis of leading MicroBanks. Econ J 117(517):107–133

Fernando NA (2006) Understanding and dealing with high interest rates on microcredit.Asian Development Bank. http://www.adb.org/documents/books/interest-rates-microcredit/microcredit-understanding-dealing.pdf. Retrieved 25 Dec 2011

Ghatak M (2000) Screening by the company you keep: joint liability lending and the peer selectioneffect. Econ J 110:601–631

Ghatak M, Guinnane TW (1999) The economics of lending with joint liability: theory and practice.J Dev Econ 60:195–228

Gibbons DS, Meehan JW (2002) Financing microfinance for poverty reduction. http://www.microcreditsummit.org/papers/financing.pdf. Retrieved 18 Dec 2010

Gonzalez-Vega C (1998) Microfinance: broader achievements and new challenges, economics andsociology. Rural finance program. Department of agricultural, environmental and developmenteconomics. The Ohio State University. Occasional paper no. 2518

Hartarska V, Nadolnyak D (2007) Do regulated microfinance institutions achieve better sustain-ability and outreach? Cross-country evidence. Appl Econ 39(10):1207–1222

Helms B, Reille X (2004) Interest rate ceilings and microfinance: the story so far. ConsultativeGroup to Assist the Poor (CGAP) Occasional paper 9, Washington, DC

Hulme D, Mosley P (1996) Finance against poverty. Routledge, LondonIntellecap (2010) Indian microfinance crisis of 2010: turf war or a battle of intentions? An

Intellecap white paper. Intellecap, HyderabadKarnani A (2011) Microfinance needs regulation. Stanford Soc Innovat Rev 9:48–53Ledgerwood J (1999) Microfinance handbook: an institutional and financial perspective. World

Bank, Washington, DCLewis JC (2008) Microloan sharks. Stanford Soc Innovat Rev, Summer:54–59Mahajan V, Nagasri G (1999) Building sustainable microfinance institutions in India. Paper

presented at session on India: the emerging microfinance market. Frankfurt Seminar on NewDevelopment Finance

Malegam Committee Report (2011) Report of the Reserve Bank of India sub-committee of itsCentral Board of Directors to Study Issues and Concerns in the Micro Finance Institutions(MFI) Sector. Reserve Bank of India. http://www.rbi.org.in/scripts/BS_PressReleaseDisplay.aspx?prid=23780. Retrieved 25 Feb 2010

Meyer RL (2002) Track record of financial institutions in assisting the poor in Asia. AsianDevelopment Bank Institute (ADB) Research Paper No. 49: Tokyo

Morduch J (1999) The microfinance promise. J Econ Lit 37(4):1569–1614Morduch J (2000) The microfinance schism. World Dev 28(4):617–629Navajas S, Schreiner M, Richard M, Claudio G, RodriguezMeza J (2000) Microcredit and the

poorest of the poor: theory and evidence from Bolivia. World Development 28(2):333–346Nyamsogoro GD (2010) Financial sustainability of rural microfinance institutions in Tanzania.

Published Doctoral Thesis, University of GreenwichOlivares-Polanco F (2005) Commercializing microfinance and deepening outreach? Empirical

evidence from Latin America. J Microfinanc 7(2):38–40Otero M (2000) Bringing development back to micro finance. J Microfinanc 1(1):8–19Otero M, Rhyne E (eds) (1994) The new world of microenterprise finance: building healthy

financial institutions for the poor. Kumarian Press, West HartfordPanwar JS (2011) Microfinance in India: mission or misery? Responsible Research, SingaporePissarides F, Nussambaumer M, Gray C (2004) Sustainability of microfinance banks: the ultimate

goal. http://www.ebrd.com/pubs/lawp/lit/03a/sustain.pdf. Retrieved 9 Mar 2005Qayyum A, Ahmad M (2006) Efficiency and sustainability of micro finance institutions in South

Asia. MPRA Paper 11674, University Library of Munich. http://www.saneinetwork.net/pdf/SANEI_VI/SANEIVI%20PROJECT%207%20Efficiency%20and%20Sustainability%20of%20Micro%20Finance%20Institutions%20in%20South%20Asia.pdf.. Retrieved 25 Dec 2010

Page 52: Sustainability of Indian Microfinance Institutions ||

36 2 Literature Review: The Sustainability Debate

Rhyne E (1994) A new view of finance program evaluation. In: Otero M, Rhyne E (eds) The newworld of microenterprise finance: building healthy financial institutions for the poor. KumarianPress, West Hartford

Rhyne E (1998) The Yin and Yang of microfinance: reaching the poor and sustainability.MicroBank Bull 2(1):6–8

Rhyne E, Guimon A (2007) The Banco Compartamos initial public offering. ACCION InsightRobinson SM (2001) The microfinance revolution: sustainable finance for the poor. World Bank,

Washington, DCRosenberg R (2007) Consultative Group to Assist the Poor (CGAP) reflections on the Compar-

tamos initial public offering: a case study on microfinance interest rates and profits: CGAPoccasional paper, Washington DC

Rosenberg R (2009) Measuring results of microfinance institutions minimum: indicators thatdonors and investors should track-a technical guide, Consultative Group to Assist the Poor(CGAP), Washington, DC

Rosenberg R, Gonzalez A, Narain S (2010) The new moneylenders: are the poor being exploitedby high microcredit interest rates? In: Todd AW, Karen H (eds) Moving beyond storytelling:emerging research in microfinance (Contemporary studies in economic and financial analysis),vol 92. Emerald Group Publishing Limited, United Kingdom, pp 145–181

Ross A, Savanti P (2005) Empirical analysis of the mechanisms of group lending. Centerfor microfinance research working paper series. http://ifmr.ac.in/cmf/publications/wp/2005/7_ross_paula-empricalanalysis.pdf. Retrieved 6 May 2010

Samarapally A, Gaul S (2011) Reviewing the Reserve Bank of India’s microfinance framework.Microfinance Information Exchange Publications

Savita S (2007) Transaction cost in Group Microcredit in India. Case studies of three micro financeinstitutions. Center for microfinance research working paper series. http://www.ifmr.ac.in/cmf/publications/wp/2006/13_shankar-caseStudyMFIs.pdf. Retrieved 6 May 2010

Scholtens B, Wensveen DV (2003) The theory of financial intermediation: an essay on what it does(not) explain. In: Balling M (ed) IDEAS: http://www.suerf.org/download/studies/study20031.pdf. Retrieved 24 June 2010

Schreiner M (1996) Thinking about the performance and sustainability of microfinance organiza-tions. http://citeseerx.ist.psu.edu/viewdoc/summary. Retrieved 9 May 2010

Sharma SR, Nepal V (1997) Strengthening of credit institutions/programs for rural povertyalleviation in Nepal. United Nations, Economic and Social Council (ECOSOC) for Asia andPacific. Bangkok

Simanowitz A (2002) Ensuring impact: reaching the poorest while building financially self-sufficient institutions, and showing improvement in the lives of the poorest women and theirfamilies. In: Daley-Harris S (ed) Pathways out of poverty: innovations in microfinance for thepoorest families. Kumarian Press, Bloomfield

Sinha S (2010) How to calm the charging bull: an agenda for CGAP in the decade of the Teneeis.Micro-Credit Rating International Limited, Gurgoan

Stiglitz JE, Weiss A (1981) Credit rationing in markets with imperfect information. Am Econ Rev71:393–410

Swami P, Shekar M, Choksey N (2010) Ruffled feathers. Bus India 11:58–66Thapa G (2007) Sustainability and governance of microfinance institutions: recent experiences and

some lessons for Southeast Asia. Asian J Agricul Dev 4(1):17–37Thorat YSP (2006) Microfinance in India: sectoral issues and challenges. Towards a sustainable

microfinance outreach in India. NABARD, GTZ and SDC, New Delhi, pp 27–42Von Pischke J (1996) Measuring the trade-off between outreach and sustainability of microen-

treprise lenders. J Int Dev 8:225–239Woller G (2000) Reassessing the financial viability of village banking: past performance and future

prospects. Microbank Bull. Microfinance Information Exchange, 5:3–8Woller G, Schreiner M (2002) Poverty lending, financial self-sustainability and the six aspects of

outreach. The SEEP network. Poverty Lending Working Group, Washington, DC

Page 53: Sustainability of Indian Microfinance Institutions ||

References 37

Woller G, Woodworth W (2001) Microcredit as a grass-roots policy for international development.Policy Stud J 29(2):267

Woller G, Dunford C, Warner W (1999) Where to microfinance. Int J Econ Dev 1:29–64Yaron J (1992) Successful rural finance institutions. World Bank discussion paper no. 150. World

Bank, Washington, DCYunus M (2009) Lifting people worldwide out of poverty. Knowledge@Wharton

Page 54: Sustainability of Indian Microfinance Institutions ||

Chapter 3Research Objectives and Design

3.1 Preface

The previous two chapters were devoted to discuss the research problem pursuedin this study. By introducing the concept of microfinance and reviewing literaturein the realm of microfinance sustainability, the research gaps and proposed researchobjectives were discussed. This chapter takes this proposal forward by framing theresearch objectives, in the light of the research design formulated in this study. Athree-phased sequential explanatory mixed-methods research design is formulatedin this study to accomplish the research objectives. This mixed-methods researchdesign along with the paradigm elements that underpin it is discussed in this chapter.

3.2 Research Objectives and the Sequential ExplanatoryMixed-Methods Research Design

This section revisits the research objectives pursued in this study within theframework of the research design formulated. A schematic representation of thethree-phased sequential explanatory mixed-methods research design formulated forthis study is given below in Table 3.1.

Table 3.1 shows that a mixed-methods research design is used in this study. In thewords of Creswell and Clark (2007), ‘Mixed Methods Research is a research designwith philosophical assumptions as well as methods of inquiry. As a methodology,it involves philosophical assumptions that guide the direction of the collection andanalysis of data and the mixture of qualitative and quantitative approaches in manyphases in the research process. As a method; it focuses on collecting, analyzing,and mixing both quantitative and qualitative data in a single study or a series ofstudies. Its central premise is that the use of quantitative and qualitative approachesin combination provides a better understanding of research problems than eitherapproach alone’.

N. Marakkath, Sustainability of Indian Microfinance Institutions:A Mixed Methods Approach, India Studies in Business and Economics,DOI 10.1007/978-81-322-1629-2__3, © Springer India 2014

39

Page 55: Sustainability of Indian Microfinance Institutions ||

40 3 Research Objectives and Design

Tabl

e3.

1R

esea

rch

desi

gn

Sequ

entia

lexp

lana

tory

mix

ed-m

etho

dsre

sear

chde

sign

Phas

eI

Phas

eII

Phas

eII

I

Qua

ntita

tive

phas

eIn

term

edia

teph

ase

Qua

litat

ive

phas

e

Res

earc

hob

ject

ives

Lite

ratu

rere

view

The

oret

ical

mod

elR

egre

ssio

nan

alys

isD

iscr

imin

ant

anal

ysis

Dat

aen

velo

pmen

tan

alys

isIn

terv

iew

ing

I.To

iden

tify

the

fact

ors

affe

ctin

gO

SSof

Indi

anM

FIs

**

*

II.T

oid

entif

yth

efa

ctor

sth

atdi

scri

min

ate

the

OSS

stat

usof

Indi

anM

FIs

*

III.

Toid

entif

yth

eef

ficie

ntan

dsu

stai

nabl

eIn

dian

MFI

s

*

IV.T

oun

ders

tand

abou

tthe

man

agem

ento

fth

efa

ctor

saf

fect

ing

and

disc

rim

inat

ing

the

OSS

stat

usof

Indi

anM

FIs

**

Not

e:T

heas

teri

skm

ark

indi

cate

sth

ere

spec

tive

rese

arch

met

hods

used

for

fulfi

lling

the

rese

arch

obje

ctiv

es

Page 56: Sustainability of Indian Microfinance Institutions ||

3.2 Research Objectives and the Sequential Explanatory Mixed-Methods Research Design 41

A sequential explanatory mixed-methods research design is framed for this studyas a combination of both quantitative and qualitative inquiry is considered essentialto address the research problem in a comprehensive manner (Green et al. 1989;Tashakkori and Teddlie 1998). The rationale for using mixed-methods frameworkis because neither quantitative nor qualitative methods are sufficient by themselvesto capture and explain the complexity of this research problem. The rationale forusing a three-phased mixed-methods sequential explanatory design to address thisresearch problem is as follows:

(a) Phase I—Quantitative phase: In this phase the factors affecting and discriminat-ing the OSS status of Indian MFIs will be identified by collecting numerical dataon the MFIs. The quantitative data analysis and results of this phase will help theresearcher to isolate and focus on the factors which needs further investigationin this research work. This phase is pursued with a multiple regression analysisand a multiple discriminant analysis to fulfil the first two objectives of thisresearch work. A regression analysis is undertaken to identify the factorsaffecting the OSS of Indian MFIs. A theoretical model of the factors affectingthe sustainability of MFIs is framed, on the basis of literature review. Basedon this model, hypotheses are formulated that denote the expected relationshipof the factors with the OSS of an MFI. Using multiple regression analysisthese hypotheses are tested on a sample of 50 Indian MFIs. This analysisthus accomplishes the first objective of this study by identifying the significantfactors affecting the OSS of Indian MFIs. Subsequent to the accomplishmentof the first objective, a multiple discriminant analysis is undertaken to fulfilthe second objective of identifying the factors that discriminate or predict theOSS status of Indian MFIs. Thus, together the two analysis works undertakenin the quantitative phase, identifies the factors that Indian MFI managers shouldconcentrate on to enhance the OSS of their MFIs.

(b) Phase II—Intermediate Phase: In this phase, the researcher will identify theparticipant MFIs who can be followed up with interviews, to understandabout the managerial aspects of the factors identified in the quantitative phase.Therefore, this phase is pursued as an intermediate participant selection phase.This phase acts as a link that connects the quantitative phase of the study to itssubsequent qualitative phase by adopting a participant selection strategy. Here,from among the 50 MFIs, Indian MFIs which are efficient and sustainable areidentified. MFIs which have relatively high levels of efficiency are identifiedusing a non-parametric method called Data Envelopment Analysis (DEA) andtheir sustainability is assessed by framing a ‘Sustainability Diamond Model’.Thus, the analysis and results of this phase identifies a set of efficient andsustainable Indian MFIs. This fulfils the third objective of the study. Out ofthese identified MFIs, those willing to cooperate with this study are regardedas participants for the ensuing phase of the research. These efficient andsustainable participant MFIs are then subject to a qualitative inquiry, in orderto understand how they are managing the factors identified in the quantitativephase of the study.

Page 57: Sustainability of Indian Microfinance Institutions ||

42 3 Research Objectives and Design

Multiple RegressionAnalysis & Multiple

DiscriminantAnalysis

Data EnvelopmentAnalysis

(To Identify Efficient &Sustainable Indian MFIs)

Interviewing & Literature Review

Factors Affecting &Discriminating the

OSS of Indian MFIs

Strategies Used forManaging the

Identified Factors &Issues in their

Mismanagement

Quantitative Phase IntermediateParticipant Selection

Phase

Qualitative Phase

Fig. 3.1 Sequential phases in the mixed-methods research design

(c) Phase III—Qualitative Phase: In this phase a qualitative analysis is undertakento follow-up the participant MFIs selected in the intermediary phase. A follow-up qualitative analysis in the nature of interviewing is conducted on the MFImanagers of selected Indian MFIs that are efficient and sustainable in their oper-ations. Qualitative data is collected from practitioners, so as to understand howthey would explain the relationships and management of the factors identifiedin the quantitative phase. Confirmation is sought from these managers aboutthe relationship of the factors identified quantitative phase of the study. Theyare also asked to explain how they are managing these significant determinantand discriminant factors. This phase is then concluded by drawing referencefrom Indian microfinance crisis and its associated literature, to understand theissues involved in the mismanagement of these determinant and discriminantfactors. Together it completes the fourth objective of the study, which aims tounderstand the managerial issues.

Thus, the three phases together fulfils the research aim of understanding theissues related to managing the sustainability of Indian MFIs. Together the threephases use an explanatory approach to fulfil this research aim. The sequential phasesinvolved in this mixed-methods research is depicted below in Fig. 3.1.

3.3 Paradigm Elements in Mixed-Methods Research Design

As discussed above, the study uses a research design that mixes both quantitativeand qualitative research methods. By doing so it adheres to the philosophy of mixed-methods research called ‘pragmatism’. Pragmatism is the philosophy where the

Page 58: Sustainability of Indian Microfinance Institutions ||

3.3 Paradigm Elements in Mixed-Methods Research Design 43

Table 3.2 Paradigm elements in a mixed-methods research design (Source: Creswell and Clark2007)

Paradigm elementsQuantitative phasepostpositivism

Qualitative phaseconstructivism

Mixed-methods studypragmatism

Ontology (what isnature ofreality?)

Singular reality (accept orreject the hypothesisformulated on thebasis of a prioritheoretical framework)

Multiple realities(illustrates differentperspectives of theproblem under studythrough participantquotes)

Singular and multiplerealities (testhypothesis and alsobuild on it throughmultipleperspectives)

Epistemology(what is therelationshipbetween theresearcher andthat beingresearched?)

Distance and impartiality(objectively collectdata from financialstatements)

Closeness (visitsparticipants tocollect data andgains an emicperspective of theproblem)

Practicality (collectsdata on the basis of‘what work’ toaddress the researchquestion)

Axiology (what isthe role ofvalues)

Unbiased (uses checks toeliminate bias)

Biased (explicitly statesabout probablebiases that havecrept ininterpretation)

Multiple stances(includes bothbiased and unbiasedperspectives)

Methodology(what is theprocess ofresearch?)

Deductive (tests an apriori theoreticalframework based onliterature)

Inductive (starts withparticipants views tobuilds up patterns,common factors,codes and themes)

Combining (collect bothqualitative andquantitative data andmixes them)

Rhetoric (what isthe language ofresearch?)

Formal style (usesstandardizeddefinitions forvariables)

Informal style(inferences frominterviews arewritten in a literaryinformal style)

Formal and informalstyles (employs bothformal and informalstyles of writing toaddress the researchproblem)

researcher’s focus is on the research objective and the consequences of research(Creswell and Clark 2007). It is pluralistic and oriented towards ‘what works’from a practical perspective. It combines research methods to gain pragmaticview of the research problem. Though the overall philosophy or paradigm of thiswork is pragmatism, individually it adheres to the philosophy of postpositivism,in the quantitative phase and constructivism, in the qualitative phase. The paradigmelements in these phases are discussed below in Table 3.2. The paradigms have beensketched based on the philosophical foundations outlined by Creswell and Clark(2007) in their work on mixed-methods research.

The research study thus uses a mixed-methods framework that pragmaticallyreaps the synergies of combining both quantitative and qualitative methods ofinquiry. In the quantitative phase the researcher advances knowledge, by using apostpositivist approach for testing hypotheses and conducting a cause and effectanalysis. As a result of this deductive approach, the researcher narrows down to

Page 59: Sustainability of Indian Microfinance Institutions ||

44 3 Research Objectives and Design

specific variables, i.e. the determinants and discriminants of OSS that needs furtherinvestigation. The qualitative phase complements the findings of the quantitativephase, by further advancing knowledge through the use a constructivist or partici-patory approach. In the qualitative phase, data is collected from the practitioners ofefficient and sustainable MFIs, as to how they are managing the factors identifiedin the quantitative phase. As a result of this inductive approach, the researchergains a holistic understanding of the strategies used by efficient and sustainableMFIs to manage their OSS. A discussion on the mismanagement of OSS is alsopresented in this phase. Thus, by combining these two approaches in a mixed-methods research framework, the study adheres to the paradigm of pragmatism.Pragmatism is a paradigm that believes in the compatibility of quantitative andqualitative methods in holistically addressing a research problem. Thus in a mixed-methods study, the researcher builds knowledge by pragmatically asserting truth as‘what works’ (Howe 1988).

3.4 Summary

This chapter portrays the research objectives of this study in the context of itsproposed research design. The rationale for formulating a three-phased sequentialexplanatory mixed-methods research design is explained in this chapter. The under-lying philosophies and synergies of combining both quantitative and qualitativeresearch methods in a mixed-methods research design are also elucidated. The sub-sequent chapters will unveil how this mixed-methods research design accomplishesthe objectives of this study, by collecting and analysing numeric data and text datain a sequential manner.

References

Creswell JW, Plano Clark VL (2007) Designing and conducting mixed methods research. Sage,Thousand Oaks

Greene JC, Caracelli VJ, Graham WF (1989) Towards a conceptual framework for mixed-methodevaluation designs. Edu Eval Pol Anal 11:255–274

Howe KR (1988) Against the quantitative-qualitative incompatibility thesis or Dogmas die hard.Edu Research 17:10–16

Tashakkori A, Teddlie C (1998) Mixed methodology: combining qualitative and quantitativeapproaches. Sage, Thousand Oaks

Page 60: Sustainability of Indian Microfinance Institutions ||

Chapter 4Quantitative Phase: Identification of FactorsAffecting and Discriminating Sustainability

4.1 Preface

The previous chapter portrayed the four objectives of the study in the broadframework of its three-phased sequential explanatory mixed-methods researchdesign. Against this backdrop, in this chapter the first phase of the study is pursued.The first phase is quantitative in nature. This phase attempts to fulfil the first andsecond objectives of the study, being (1) identification of factors affecting the OSSof Indian MFIs and (2) identification of the factors discriminating the OSS status ofIndian MFIs. An elaborate discussion on this quantitative phase of this study, whichidentifies the determinants and discriminants of OSS, is presented in this chapter.

4.2 Structure of the Quantitative Phase

The structure of the processes involved in the quantitative phase is portrayed inFig. 4.1. For achieving the first objective, a literature review is undertaken on thefactors affecting the OSS of MFIs. A theoretical model is formulated based onthis literature support. This theoretical model is subsequently used to formulatehypotheses which depict the expected relationships that these factors share withthe OSS of MFIs. The hypotheses are then tested on a sample of Indian MFIs,using multiple regression technique. Data is sourced from 50 Indian MFIs forthis purpose. The results of this regression analysis identify the significant factors

This chapter has material from the article: What discriminates the operational self-sustainabilityof Indian MFIs: A multiple discriminant analysis inquiry. Marakkath N, Ramanan TR (2012).Paper presented at 3rd international conference on Institutional and Technological Environmentfor Microfinance (ITEM3) on Cost 650 Management & Social Performance in Microfinance, NewDelhi, 4–7 Jan 2012. Organized by Burgundy School of Business, France (Secured Top PaperPresentation Award)

N. Marakkath, Sustainability of Indian Microfinance Institutions:A Mixed Methods Approach, India Studies in Business and Economics,DOI 10.1007/978-81-322-1629-2__4, © Springer India 2014

45

Page 61: Sustainability of Indian Microfinance Institutions ||

46 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

Quantitative Phase

Multiple Regression Analysis Multiple Discriminant Analysis

Factors Affecting OSS Statusof Indian MFIs

Factors Discriminating OSSStatus of Indian MFIs

Objective 1: To identify thefactors affecting the OSS ofIndian MFIs

Objective 2: To identify thefactors discriminating theOSS status of Indian MFIs

Literature Review

Data & Model

Fig. 4.1 Structure of quantitative phase

affecting the OSS of Indian MFIs. Thereafter the second objective of identifyingthe discriminants of Indian MFI’s OSS status is pursued. The significant factorsidentified as determinants of OSS and two component variables of OSS ratioare regarded as probable discriminators of the OSS status of Indian MFIs. Thepredictive power of these identified probable discriminators is tested and validatedusing multiple discriminant analysis model. The results of this discriminant analysisidentify the significant factors that are capable of discriminating or predicting theOSS status of Indian MFIs.

The processes sketched in Fig. 4.1 are elucidated in the ensuing sections of thischapter.

4.3 Literature Review on the Factors Affectingthe Operational Self-Sustainability of MicrofinanceInstitutions

A literature review on the factors influencing the OSS of MFIs was undertaken.Based on this literature survey, the factors that influence the OSS of MFIs areclassified into five broad categories. The five categories are portfolio risk factor,capital structure factor, development factor, growth factor and institutional factors.Each of these factors is represented in this work, by selecting proxy variables.These proxy variables are used to formulate a theoretical model which hypothesizesthe expected relationships of these variables with the OSS ratio of MFIs. Thehypotheses are then tested on a sample of Indian MFIs, by using these proxy

Page 62: Sustainability of Indian Microfinance Institutions ||

4.3 Literature Review on the Factors Affecting the Operational Self-Sustainability. . . 47

variables as exogenous variables in a multiple regression model. This section isa discussion on literature review undertaken to identify these factors and proxyvariables.

4.3.1 Portfolio Risk Factor

This factor denotes the quality of MFI’s loans. Cull et al. (2007) and Ghatak andGuinnane (1999) note that sustainable MFIs maintain the quality of its loan portfolioby disbursing group loans, with joint liability on all the group members. Peerpressure and threat of social punishment within the groups effectively replaces theneed for physical collateral and ensures high recovery rates for MFIs. Repaymentrate and efficiency is seen higher under joint-liability contracts as compared toconventional individual-liability contracts because the former exploits a usefulresource that the latter does not—the information borrowers have about each otherin the groups (Ghatak 2000). This reduces the information asymmetric credit marketrisks in lending operations. Though this has been the experience in India, themicrofinance crisis in the district of Andhra Pradesh has deteriorated the portfolioquality of Indian MFIs. Uncontrollable metrics of portfolio quality like portfolio atrisk greater than 30 days and recovery rates were found to be adversely affecteddue to the crisis (Intellecap 2010). This has implications for the sustainability ofIndian MFIs. Therefore in tune with past works by Ayayi and Sene (2007) andCrombrugghe et al. (2008), this work uses portfolio at risk greater than 30 daysratio as a proxy variable to capture the portfolio riskiness of an MFI.

4.3.2 Capital Structure Factor

This factor denotes the structure of an MFI’s capital mix. The impact of capitalstructure factors on the OSS of MFIs have been studied by Coleman (2007) andBogan (2008). Coleman (2007) studies the impact of leveraged capital structure onthe sustainability of MFIs and reports a positive relationship between the debt andsustainability. Bogan (2008) confirms the same finding with respect to debt, butreports a negative association between donations and financial self-sustainabilityof MFIs. In tune with these findings, this work uses two proxy variables tocapture the effect of capital structure on operational self-sustainability—equity toassets ratio and donation to assets ratio [Equity to assets ratio D (Equity C Retainedearnings)/Total assets] and [Donations to assets ratio D Donations/Total assets].

4.3.3 Development Factor

This factor denotes the development orientation of an MFI or depth of an MFI’soutreach (i.e. ability of MFI in reaching out to the very poor clientele). This can

Page 63: Sustainability of Indian Microfinance Institutions ||

48 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

be captured by poverty level and gender of the clients (Christen 2001; Navajaset al. 2000; Bhatt and Tang 2001; Olivares-Polanco 2005; Von Pischke 1996). Theassumptions in these studies are that the greater the number of poor clientele andwomen clientele served by the MFI, the deeper is the outreach. These studiesperceive average loan size per borrower of the MFI to be a proxy for povertylevel of clientele and regard women clientele to be poorer than men. They alsodeliberate on the trend seen among MFIs, to adopt commercialized managerialpractices to remain sustainable, thereby drifting from the mission of serving thepoor. This discussion makes it interesting to study the relationship between an MFI’sOSS and the mission drift issue faced by it. Going by this observation, this workintroduces average loan size per borrower, [Average loan size per borrower D Grossloan portfolio/Number of borrowers] a proxy variable for the poverty level ofclientele, to see if there is a trade-off effect of mission drift for MFIs, while pursuingthe goal of sustainability. The impact of serving the women clientele [Womenborrowers D Number of women borrowers] on the OSS of MFIs is also studiedin this work. D’Espallier et al. (2011) observe that the presence of more womenclientele for an MFI is associated with lower portfolio at risk, lower write-offs andlower credit-loss provisions, all leading to higher OSS.

4.3.4 Growth Factor

This factor denotes the scale of MFI’s operations. Scale is vital for an MFI to achieveits OSS (Nisha 2007). Qayyum and Ahmad (2006) observe economies of scale todirectly influence sustainability of MFIs in India. Crombrugghe et al. (2008) test theimpact of growth on the sustainability of 42 Indian MFIs, using gross loan portfolioand total number of borrowers as proxies for growth. Similarly, Ayayi and Sene(2007) test the influence of growth on the sustainability of a sample of 217 MFIsin 101 countries, using client outreach as a proxy for growth. The results of boththe studies confirm the positive influence that growth has on sustainability of MFIs.In similar lines, Nair (2005) also suggests that scale economies could be reaped byIndian MFIs by pursuing growth. In tune with the observations of these prior works,this work also hypothesizes a positive relationship between growth and OSS, usinggross loan portfolio as proxy variable for growth.

4.3.5 Institutional Factor

This factor denotes the aspects specific to an MFI, which affects its OSS. Prominentvariables in this category are discussed in literature by Venkatraman and RajSekhar(2008), Ayayi and Sene (2007) and Crombrugghe et al. (2008). Venkatraman andRajSekhar (2008) in their study note MFIs which are regulated in nature to bewell governed compared to their unregulated counterparts and to have higher levels

Page 64: Sustainability of Indian Microfinance Institutions ||

4.4 Data, Theoretical Model and Hypotheses 49

of sustainability in India. Ayayi and Sene (2007) in their study hypothesize ageas a variable sharing a direct relationship with sustainability. Apart from age,Crombrugghe et al. (2008) in their study denote the location of MFI, credit deliverymodel used by MFI and savings facilities provided by MFIs to have an influenceon an MFI’s sustainability. Going by the findings of these studies, this workincorporates manageable institution-specific variables like location, credit deliverymodel used by MFI, savings facilities provided by MFIs and regulatory status ofMFIs to the regression model. MFIs which use the home-grown Self-Help Group(SHG) model and Bangladeshi Grameen model are differentiated using dummycoding. As there were only two MFIs in the sample which has less than 5 % oftheir lending in the form of individual lending, a separate category for individuallending is not maintained. Age, an uncontrollable institutional variable, is alsotested in order to study its impact on OSS. Governance of MFIs is not capturedas a separate variable, as there is no data available on this aspect for all the sampleMFIs. Though it can be expected to be partially captured by regulatory status of theMFI, as observed by Venkatraman and RajSekhar (2008), it is not a proxy capable ofcapturing the nuances of MFI governance. Other uncontrollable factors pertaining tothe macroeconomic conditions of an MFI are not tested in this work, as the intentionis to arrive at a set of factors that influences an MFI’s OSS, which are manageableby the MFI, if not wholly but at least partially.

Thus, with literature support, the variables that proxy each of the five factors areidentified. The impact of change in each of these variables is studied over changein the OSS of these MFIs. The change window is taken as 2005–2009. The datasourced for testing the significance of these factors and the theoretical model coinedto formulate the hypotheses on these factors are discussed in the next section.

4.4 Data, Theoretical Model and Hypotheses

After reviewing the factors that influence OSS, the data on which the significanceof these factors can be tested is sourced. Data for this phase is sourced from MIXMarket Database. As discussed in the introductory chapter, there is paucity of dataon Indian MFIs. This makes it difficult to conduct a large sample study. Afterreviewing the MIX Market Database, 50 Indian MFIs were found to have disclosedtheir standardized operational data for the period 2005–2009. This work is thuslimited to these 50 MFIs which have disclosed their data to MIX for the years 2005–2009. These MFIs can be regarded to be transparent and comparatively sustainablein their operations than the vast majority of MFIs that are portrayed by NABARD tobe opaque and unproven in terms of its sustainability. Data is collected on these 50MFIs. As these 50 MFIs have not consistently reported their data on all the selectedfactors for the consecutive years 2005–2009, a panel data analysis could not beundertaken. This forced the quantitative analysis to be cross-sectional in nature. Butthe longitudinal effect is captured to the possible extent, by using the change infactors for the window 2005–2009. Change in the values of these factors, over the

Page 65: Sustainability of Indian Microfinance Institutions ||

50 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

PortfolioRisk Factor

CapitalStructureFactor

DevelopmentFactor

GrowthFactor

InstitutionalFactor

Portfolio at RiskGreater than 30

Days Ratio

Donations toAssets Ratio,

Equity to AssetsRatio

Average LoanSize, WomenBorrowers

Gross LoanPortfolio

Age, Location,Credit Model,Regulatory

Status, FinancialIntermediation

Ratio

OperationalSelf-

SustainabilityRatio

Fig. 4.2 Theoretical model of factors affecting the operational self-sustainability of microfinanceinstitutions (Source: Author’s construct based on literature review)

year 2005 to the year 2009, is ascertained and used for testing the hypotheses on asample of 50 Indian MFIs.

The 50 sample MFIs used in this study are spread across the geography of thenation. Sixty-two per cent of MFIs in the sample were in South India and 38 % inNorth India. This proportion in the sample conforms to the increased concentrationof MFIs in South India, which is a representative of the MFI population inIndia.

After sourcing the data, a theoretical model is coined to formulate hypothesesthat denote the expected relationships between the factors and the OSS ratio. Themodel is depicted in Fig. 4.2.

As the model coined is a dependence model showing expected relationshipsbetween one dependent variable and several independent variables, multiple regres-sion analysis is used to test its significance. Each of the factors in the model iscaptured using variables, sourced from the balance sheets, profit and loss accountsand websites of the sample MFIs. More details on these variables are explained inthe subsequent subsections.

Page 66: Sustainability of Indian Microfinance Institutions ||

4.5 Multiple Regression Analysis and Results 51

4.4.1 Dependent Variable

For the analysis purpose, OSS ratio is used as the dependent or endogenous variable.OSS ratio is the ratio of operating income over the total cost of an MFI (i.e. operatingcosts C financing costs C loan loss provisions). A ratio above 100 % denotes thatMFI has enough operating income to cover its costs, indicating an operationallyself-sustainable status.

4.4.2 Independent Variables and Hypotheses

The 11 variables discussed earlier in Sect. 4.3 of this chapter are used as independentor exogenous variables to proxy the factors affecting OSS of MFIs. They areportfolio at risk greater than 30 days ratio, donation to assets ratio, equity to assetsratio, average loan size, number of women borrowers, gross loan portfolio, age,location, credit model, regulatory status and financial intermediation ratio.

The details on these independent variables and the hypotheses used in the studyare discussed in Table 4.1.

The hypotheses discussed above are tested using multiple regression analysis.

4.5 Multiple Regression Analysis and Results

The empirical multiple regression model tested in the phase is as follows:

� OSS D ˇ0 C ˇ1� PAR > 30 C ˇ2 � DAR C ˇ3 � EAR C ˇ4 � ALSPB

C ˇ5 � WB C ˇ6 �GLP C ˇ7 AGE C ˇ8 LOC C ˇ9 CM C ˇ10 RS

C ˇ11 � FIR C � (4.1)

where � is the change in value of the variable from 2005 to 2009, ˇ 0 is the intercept,ˇ 1 to ˇ 11 are the beta coefficients of the independent variables and � is the randomerror term.

The descriptive statistics of the metric variables used in this model are summa-rized in Table 4.2.

The descriptive statistics in Table 4.2 depict that the OSS ratio of sample MFIshas shown a positive growth for the period 2005–2009. The mean OSS for 2005and 2009 was also calculated. For the sample, the mean OSS for 2005 and 2009is 104.14 % and 117.06 %, respectively. The minimum OSS for 2005 and 2009 is4.50 % and 25.02 %, respectively, and the maximum OSS for 2005 and 2009 is

Page 67: Sustainability of Indian Microfinance Institutions ||

52 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

Tabl

e4.

1In

depe

nden

tvar

iabl

esan

dhy

poth

eses

Inde

pend

entv

aria

bles

Fact

ors

Prox

ies

Hyp

othe

ses

Fact

or1:

port

folio

risk

fact

orPo

rtfo

lio

atri

sk>

30da

ysra

tio

(PA

R>

30)D

Port

folio

atri

sk>

30da

ys/g

ross

loan

port

folio

Itde

note

sth

epo

rtfo

liori

skin

ess

ofth

eM

FIby

capt

urin

gth

eva

lue

ofal

llo

ans

outs

tand

ing

that

have

one

orm

ore

inst

alm

ents

ofpr

inci

palp

astd

uem

ore

than

30da

ysH

ypot

hesi

s1:

Cha

nge

inpo

rtfo

lioat

risk

grea

ter

than

30da

ysra

tiois

inve

rsel

yre

late

dto

chan

gein

OSS

.The

low

erth

era

tio,t

hebe

tter

the

qual

ityof

loan

port

folio

and

the

high

erth

eO

SSof

the

MFI

Fact

or2:

capi

tals

truc

ture

fact

orD

onat

ions

toas

sets

rati

o(D

AR

)D

Don

atio

ns/to

tala

sset

sT

his

deno

tes

the

dona

tions

used

byan

MFI

inits

capi

tals

truc

ture

Hyp

othe

sis

2:C

hang

ein

dona

tion

toas

sets

ratio

isdi

rect

lyre

late

dto

chan

gein

OSS

.Tho

ugh

ane

gativ

ere

latio

nshi

pis

shar

edby

dono

rde

pend

ency

with

finan

cial

self

-suf

ficie

ncy

ratio

,with

OSS

itha

sa

posi

tive

asso

ciat

ion.

Thi

sis

sobe

caus

e,as

the

MFI

perf

orm

sw

ella

ndat

trac

tsm

ore

dono

rfu

nds,

the

dona

tions

acta

sca

taly

stfo

ren

hanc

ing

itsO

SSE

quit

yto

asse

tsra

tio

(EA

R)

D(E

quity

Cre

tain

edea

rnin

gs)/

tota

lass

etsD

(pai

din

capi

talC

reta

ined

earn

ings

Csh

are

prem

ium

Ctr

easu

rysh

ares

Ceq

uity

rese

rves

)/to

tala

sset

s

Thi

sde

note

sow

nfu

nds

and

plou

ghed

back

cont

ribu

tions

inth

eca

pita

lst

ruct

ure

ofa

MFI

Hyp

othe

sis

3:C

hang

ein

equi

tyto

asse

tsra

tiois

inve

rsel

yre

late

dto

chan

gein

OSS

.The

reas

onis

that

from

asu

stai

nabi

lity

pers

pect

ive,

leve

rage

oneq

uity

isve

ryim

port

antf

oran

MFI

’sgr

owth

,as

scal

eca

ndi

lute

orof

fset

fixed

cost

s.L

ever

age

may

also

boos

tpro

fitab

ility

whe

nth

eco

sts

offin

anci

ngdo

note

xcee

dth

em

argi

nalr

even

uege

nera

ted

from

itFa

ctor

3:de

velo

pmen

tfa

ctor

Ave

rage

loan

size

per

borr

ower

(AL

SPB

)D

Gro

sslo

anpo

rtfo

lio/n

umbe

rof

borr

ower

sT

his

deno

tes

the

pove

rty

leve

lof

the

clie

ntel

eH

ypot

hesi

s4:

Cha

nge

inav

erag

elo

ansi

zeis

dire

ctly

rela

ted

toch

ange

inO

SS.T

hehi

gher

the

aver

age

loan

size

,the

low

erth

epo

vert

yle

velo

fth

ecl

ient

ele

and

the

bette

rth

eM

FIsu

stai

nabi

lity

Thi

sva

riab

leis

intr

oduc

edto

see

ifth

ere

isa

trad

e-of

fef

fect

ofm

issi

ondr

ift,

whi

lepu

rsui

ngsu

stai

nabi

lity.

An

incr

ease

inlo

ansi

ze,t

houg

hre

duce

str

ansa

ctio

nco

ston

loan

san

dau

gmen

tssu

stai

nabi

lity,

indi

cate

sa

drif

tfr

omth

em

issi

onof

reac

hing

the

poor

Page 68: Sustainability of Indian Microfinance Institutions ||

4.5 Multiple Regression Analysis and Results 53

Wom

enbo

rrow

ers

(WB

)DN

umbe

rof

wom

enbo

rrow

ers

Itde

note

sth

enu

mbe

rof

fem

ale

clie

ntel

e,am

ong

the

tota

lnum

ber

ofac

tive

clie

nts

ofth

eM

FI.I

nm

icro

finan

ce,t

hebe

lief

isth

atfe

mal

ecl

ient

ele

need

sto

beem

pow

ered

thro

ugh

finan

cial

stre

ngth

;for

wom

enar

epe

rcei

ved

tobe

poor

erth

anm

enan

dle

ssau

tono

mou

sin

allfi

nanc

ial

resp

ects

Hyp

othe

sis

5:C

hang

ein

num

ber

ofw

omen

borr

ower

sis

dire

ctly

rela

ted

toch

ange

inO

SS.T

his

isso

asin

mic

rofin

ance

wom

enar

em

ore

repu

ted

for

repa

ymen

tsth

anm

enFa

ctor

4:gr

owth

fact

orG

ross

loan

port

foli

o(G

LP

)T

his

deno

tes

the

outr

each

orsc

ale

inlo

andi

sbur

sem

ents

achi

eved

byth

eM

FI,w

hich

intu

rnin

dica

tes

itsgr

owth

Hyp

othe

sis

6:C

hang

ein

gros

slo

anpo

rtfo

liois

dire

ctly

rela

ted

toch

ange

inO

SS.T

hehi

gher

the

gros

slo

anpo

rtfo

lioof

the

MFI

,the

bette

ris

the

OSS

ofth

eM

FIFa

ctor

5:in

stitu

tiona

lfac

tor

Age

(AG

E)

Itde

note

sth

enu

mbe

rof

year

ssi

nce

ince

ptio

nof

the

MFI

Hyp

othe

sis

7:A

geof

MFI

isdi

rect

lyre

late

dto

the

chan

gein

the

OSS

.Age

isex

pect

edto

have

apo

sitiv

ere

latio

nshi

pw

ithM

FIsu

stai

nabi

lity,

asth

eM

FIis

expe

cted

tobe

have

mat

ured

and

gain

edex

peri

ence

ines

tabl

ishi

ngits

oper

atio

nsw

ithag

eL

ocat

ion

(LO

C)

Itde

note

sth

ege

ogra

phic

allo

catio

nof

the

MFI

.Non

eof

the

MFI

sin

the

sam

ple

has

chan

ged

itsge

ogra

phy

ofop

erat

ion

duri

ng20

05–2

009.

Soth

ech

ange

inlo

catio

nis

notm

easu

red.

Inst

ead,

dum

my

vari

able

sar

eus

edto

dist

ingu

ish

betw

een

the

nort

hern

and

sout

hern

regi

ons,

inw

hich

the

MFI

sar

elo

cate

d.W

eus

e1

for

deno

ting

sout

hern

and

0fo

rde

notin

gno

rthe

rnre

gion

sIn

Indi

aab

ove

80%

ofth

eM

FIs

are

oper

atin

gin

Sout

hIn

dia1

.As

Sout

hIn

dia

isth

epr

efer

red

loca

tion

for

MFI

san

dha

sm

ore

liter

ate

popu

latio

n,it

isin

tere

stin

gto

stud

yif

ther

eis

adi

rect

rela

tions

hip

betw

een

pres

ence

ofM

FIin

Sout

hIn

dia

and

chan

gein

OSS

Hyp

othe

sis

8:Pr

esen

ceof

MFI

inSo

uth

Indi

ais

dire

ctly

rela

ted

toch

ange

sin

OSS

(con

tinue

d)

Page 69: Sustainability of Indian Microfinance Institutions ||

54 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

Tabl

e4.

1(c

ontin

ued)

Inde

pend

entv

aria

bles

Fact

ors

Prox

ies

Hyp

othe

ses

Cre

ditm

odel

(CM

)It

deno

tes

the

cred

itde

liver

ym

odel

used

byth

eM

FIN

one

ofth

eM

FIs

inth

esa

mpl

eha

sch

ange

dits

cred

itde

liver

ym

odel

duri

ngth

epe

riod

2005

–200

9.So

chan

gein

cred

itm

odel

isno

tmea

sure

d.In

stea

d,du

mm

yva

riab

les

are

used

todi

stin

guis

hbe

twee

nth

etw

ow

idel

yus

edcr

edit

deliv

ery

mod

els

inIn

dia—

Self

-Hel

pG

roup

(SH

G)

and

Gra

mee

nm

odel

s.W

eus

e1

for

deno

ting

SHG

and

0fo

rde

notin

gG

ram

een

mod

elIn

Indi

a,SH

Gis

the

hom

e-gr

own

mod

elfo

rm

icro

finan

ceop

erat

ions

and

Gra

mee

nis

anad

opte

dm

odel

from

Ban

glad

esh.

Soth

ein

tere

stis

tokn

owif

ther

eis

adi

rect

rela

tions

hip

betw

een

usag

eof

hom

e-gr

own

SHG

mod

elan

dth

eO

SSof

anM

FIH

ypot

hesi

s9:

Usa

geof

SHG

mod

elby

MFI

sis

dire

ctly

rela

ted

toch

ange

sin

OSS

Reg

ulat

ory

stat

us(R

S)It

deno

tes

the

lega

lsta

tus

ofth

eM

FI.W

eigh

tage

(W)

isgi

ven

toth

eM

FIs

for

the

peri

odth

eyre

mai

ned

regu

late

d.W

D0

for

MFI

sun

regu

late

dfo

rth

epe

riod

2005

–200

9,W

D5

for

MFI

sre

gula

ted

for

the

peri

od20

05–2

009

and

Wva

lue

betw

een

0an

d5

isas

sign

edto

allo

ther

MFI

sba

sed

onth

etim

epe

riod

for

whi

chth

eyre

mai

ned

regu

late

ddu

ring

the

peri

od20

05–2

009

Hyp

othe

sis

10:T

hetim

epe

riod

for

whi

chan

MFI

rem

aine

dre

gula

ted

isdi

rect

lyre

late

dto

chan

ges

inO

SS.R

egul

ated

MFI

sar

eex

pect

edto

bem

ore

tran

spar

enta

ndw

ellg

over

ned;

sour

cing

com

mer

cial

ized

fund

san

dm

obili

zing

depo

sits

,all

augm

entin

gsu

stai

nabi

lity

Fin

anci

alin

term

edia

tion

rati

o(F

IR)D

Dep

osits

/loan

sIt

deno

tes

depo

sits

mob

ilize

das

ara

tioof

loan

sdi

sbur

sed

Hyp

othe

sis

11:C

hang

ein

finan

cial

inte

rmed

iatio

nra

tiois

dire

ctly

rela

ted

toth

ech

ange

inO

SS.H

igh

finan

cial

inte

rmed

iatio

nra

tiois

expe

cted

toad

dto

the

sust

aina

bilit

yof

anM

FI,a

sth

efu

nds

from

savi

ngs

can

bepa

rked

inpr

ofita

ble

asse

tsw

hich

can

yiel

dhi

gher

rate

sof

retu

rnfo

rth

eM

FIs

1T

hedi

stri

ctof

And

hra

Prad

esh,

whe

reth

ecr

isis

onac

coun

tof

high

inte

rest

rate

sof

MFI

sw

asal

lege

d,is

also

inth

eso

uthe

rnpa

rtof

Indi

a

Page 70: Sustainability of Indian Microfinance Institutions ||

4.5 Multiple Regression Analysis and Results 55

Table 4.2 Descriptivestatistics of the metricvariables

Descriptivestatistics Mean

Standarddeviation

�OSS 10.248 37.32�EAR 5.94 24.78�DAR �3.05981 8.409723�GLP 10230.07 40360.75�WB 8591.29 27938.08�PAR > 30 �.4290 2.977�ALSPB 112.45 316.32�FIR 6.92 36.05AGE 14.54 9.65

195.05 % and 180.04 %, respectively. This shows that the sample comprises ofMFIs with varying OSS levels and is to that extent representative of the differentMFIs in the population. Similar to the positive growth in OSS ratio, the mean of allother variables, except portfolio risk ratio and donations to assets ratio, has shown apositive growth over the period 2005–2009. The negative growth in risk depicts thatover this time frame the recovery performance of the sample MFIs have improved.Similarly the negative growth in donations indicates that the MFIs are makingattempts to graduate from OSS to FSS status, which is a subsidy-independentsustainability status. Thus overall the descriptive statistics of the variables depictthat over this time frame the sample MFIs have been moving towards the goal ofsustainability.

The regression analysis was conducted on the sample data of 50 MFIs, using Sta-tistical Package for Social Science (SPSS) software version 17. The results showedthat the model is well specified with non-biased coefficients. The non-biasednessand efficiency of the coefficients were confirmed by checking for the normality andhomoscedasticity of the regression residuals. Residual normality was tested usingKolmogorov-Smirnov test and Shapiro-Wilk test. The null hypotheses for these testswere rejected, confirming normality of residuals. Homoscedasticity of residualswas tested by plotting the residuals against predicted values. The scatter plot wasrandom ensuring residual homoscedasticity. The absence of multicollinearity amongthe independent variables was ensured by checking the collinearity diagnostics—tolerance value and variance inflation factor—as depicted in Table 4.3.

Multicollinearity is a problem that arises when there is high intercorrelationamong the independent variables in the multiple regression model. This high inter-correlation makes the regression coefficients inflated and difficult for interpretation.The results presented in Table 4.3 show tolerance values above 10 and varianceinflation factor below 10. This denotes that there is no multicollinearity problemamong the independent variables included in the study (Gujarati and Sangeetha2007).

Having checked these assumptions the regression estimates are presented belowin Table 4.4.

Page 71: Sustainability of Indian Microfinance Institutions ||

56 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

Table 4.3 Collinearitydiagnostics

Collinearity statistics

Independentvariables Tolerance

Variance inflationfactor (VIF)

PAR > 30 .808 1.238DAR .850 1.177EAR .401 2.496ALSPB .216 4.640WB .266 3.761GLP .195 5.116AGE .517 1.936LOC .740 1.352CM .699 1.431RS .719 1.390FIR .671 1.490

Table 4.4 Multipleregression estimates

Hypotheses based onindependent variables Coefficient (t values)

Hypothesis 1: PAP > 30 �.297 (�2.121)*

Hypothesis 2: DAR �.109 (�.878)Hypothesis 3: EAR �.082 (�.455)Hypothesis 4: ALSPB �.587 (�2.381)*

Hypothesis 5: WB �.052 (�.235)Hypothesis 6: GLP .655 (2.513)*

Hypothesis 7: AGE �.244 (�1.535)Hypothesis 8: LOC �.198 (�1.490)Hypothesis 9: CM �.310 (�2.269)*

Hypothesis 10: RS �.175 (�1.295)Hypothesis 11: FIR .140 (1.005)Constant 49.29 (3.646)*

Adjusted R2 .359F value 3.494*

N (sample size) 50

Note: Figures in parentheses show t values*Indicates values are significant at 5 % level

The Fischer’s F test confirms the overall model fit. The F value of 3.494 which issignificant at 5 % level signifies that the model has good overall significance. Thisresult is also corroborated by the adjusted R2 of .359, which signifies that 35.90 %of the variance of the dependent variable is explained by the independent variablesin the model.2

2Compared to social science research works this is a low adjusted R square value and mayapparently give an indication that regression model has less explanatory power. But such lowadjusted R square value is typical of most financial research studies. In financial research such lowadjusted R square value is regarded quite acceptable, provided the model has overall significance inexplaining the variations in the dependent variable (i.e. if the F value of the model is significant).

Page 72: Sustainability of Indian Microfinance Institutions ||

4.5 Multiple Regression Analysis and Results 57

The regression results in Table 4.4 depict four independent variables to besignificant in the study—portfolio at risk greater than 30 days ratio, average loansize per borrower, gross loan portfolio and credit delivery model. These variablesproxy the following factors, respectively—portfolio risk factor, development factor,growth factor and institutional factor. Thus the results infer these determinants to bethe most significant factors affecting the sustainability of Indian MFIs.

Of these significant variables, except average loan size per borrower and creditmodel, both portfolio at risk greater than 30 days and gross loan portfolio compliedwith their respective hypothesized theoretical relationships. In this quantitativephase, we do not further probe on the explanations for the observed relationshipsof these significant variables. But later in the qualitative phase of this study, wewill further explore from the practitioners, confirmation on the relationship ofthese significant variables. The managers of efficient and sustainable MFIs will beasked to explain the relationships and managerial strategies associated with thesesignificant variables.

In this study though the adjusted R square value is low, the F value of the model is significant.This denotes that the independent variables used in the model have significant explanatory power.Moreover there are few specific reasons as to why the regression model used in this study has alow R square value. One of the reasons is that this study has deliberately omitted the inclusionof two determinants of sustainability—‘revenue generation factor’ and ‘cost efficiency factor’—as independent variables in the multiple regression model. Operational self-sustainability ratio(OSS ratio), the dependent variable used in the model, is nothing but a ratio of revenue and costfactors. So if we further include revenue and cost factors as independent variables, it will actas exact stand-ins or component factors for the OSS ratio. This will unduly inflate the R squarevalue and violate the statistical principle with which regression works. It will also adversely affectthe predictive power of the model (Gujarati and Sangeetha 2007). So, high R square values area matter of suspicion when it is achieved through the inclusion of exogenous variables that areexact stand-ins or components of dependent variable. So in this study we preferred to ensure thestatistical accuracy of the model rather than unduly inflating the R square value. However, theeffect of revenue generation factor and cost efficiency factor on OSS ratio was later captured usinglagged variables in the discriminant analysis model, without violating the statistical accuracy of thestudy. Secondly a statistical reason for a low R square is due to the data constraints that this studyconfronts. The sample size of the study is limited to 50 and the number of independent variablesused in the regression model is 11. In regression, the smaller the sample size and the larger thenumber of independent variables, the lower will be the adjusted R square value. This is so becausethe formula of adjusted R square is dependent on sample size and number of independent variables.

Adjusted R square D 1�((1�R square)(N�1)/(N�k�1)).

Where N D sample size and k D number of independent variables in the model.The low adjusted R square value in this study may also be attributed to the fact that the focus

of the work is limited to understanding the affect of micro-level factors on the OSS ratio of IndianMFIs. This is so as the interest of the study is to know how well the significant micro-level factorscan be managed or controlled by the MFI managers so as to enhance their MFI’s sustainability.But the fact remains that the variations in OSS ratio is also subject to the influence of severalmacroeconomic factors. The influence of these factors is not accounted in the regression modelused in the study as the affect of such uncontrollable factors does not come within the purviewof the research objectives. Thus the low adjusted R square value may also be attributed to thisselective choice of independent variables made in this study.

Page 73: Sustainability of Indian Microfinance Institutions ||

58 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

Thus in the quantitative phase, the intention of undertaking the multiple regres-sion analysis is only to identify the significant determinants of OSS. As grossloan portfolio, portfolio at risk greater than 30 days ratio, average loan size perborrower and credit delivery model are found significant, it indicates growth,portfolio risk, development and institutional factor to be the crucial factors affectingthe sustainability of MFIs. Intuitively, it infers that Indian MFI managers mustconcentrate on these factors for enhancing the sustainability of their organiza-tions.

Gross loan portfolio which denotes the MFI’s growth in loan disbursements hasthe highest beta coefficient. This indicates growth factor to be the most significantfactor contributing to OSS. Average loan size per borrower is the next significantvariable with highest beta coefficient. It has a negative relationship with OSS,contrary to the hypothesized positive relationship. Average loan size was introducedas an independent variable to see if Indian MFIs are improving their sustainabilitylevels by increasing their loan size, thereby drifting their attention to the less poorclients. But the regression results seem to show that such a mission drift does nothappen in Indian context. This would mean Indian MFIs are not profiting by driftingits mission of serving the poor. Moreover, the negative relationship shows that thepoorer the clientele, the better the sustainability. Although, this result corroboratesthe basic belief in microfinance that the poor are bankable, it also contradictsthe trade-off argument between serving the poor and attaining sustainability. Suchcontrasting results have been supported by Ashim (2010) in his dissertation onsustainability and mission drift in microfinance. Crombrugghe et al. (2008) havealso observed a similar relationship in Indian context. Credit delivery model is thenext significant variable. But contrary to the hypothesized positive relationship andthe observation made by Crombrugghe et al. (2008), the regression results depictthe SHG credit delivery model to share a negative relationship with changes inOSS. This means that MFIs with SHG model experience a negative change in OSSby �.310. Portfolio at risk greater than 30 days is the next significant variablewhich complies with the hypothesized negative relationship that is shared withOSS. The explanation for these observed relationships is investigated later in thequalitative phase of this study. Thus in this quantitative phase, by undertaking thisregression analysis, the researcher isolates and identifies the significant factors thatplay a major role in determining the sustainability of MFIs and therefore deserveconsiderable attention from MFI management.

After identifying these four significant factors, these determinant variables andthe two component factors of the OSS ratio—namely, revenue generation factorand cost-efficiency factor—are used as lagged independent variables in a multiplediscriminant analysis model to assess their predictive ability to discriminate betweenoperationally sustainable and unsustainable MFIs. Revenue generation factor andcost-efficiency factors are regarded as components of OSS ratio, because the formerdenotes the numerator and the latter denotes the denominator of the OSS ratio.Though these two component factors were also depicted in literature to haveinfluence on the OSS of MFIs, they were deliberately not included in the regressionanalysis, because it would violate a statistical principle with which regression works.

Page 74: Sustainability of Indian Microfinance Institutions ||

4.6 Selection of Probable Discriminators for the Discriminant Analysis Model 59

Regression deals with stochastic variables and not with deterministic or functionalrelationships, and therefore regressing these two component factors against OSSratio would be statistically incorrect. Including these variables would unduly inflatethe adjusted R2 of the regression model, because revenue and cost variables areverily the components of an MFI’s OSS ratio. Therefore in this study, in orderto ensure statistical accuracy, the revenue generation factor and cost-efficiencyfactor were not tested for their significance in the regression model. But these twocomponent factors are tested for their discriminatory or predictive power in thesubsequent discriminant analysis undertaken. These two factors will be includedas lagged predictor variables in the discriminant analysis model. The usage of suchlagged variables in discriminant analysis will not result in the variance inflationproblem discussed earlier.

The details of the discriminant analysis undertaken to fulfil the second objectiveof the study is discussed in the next section of this chapter.

4.6 Selection of Probable Discriminatorsfor the Discriminant Analysis Model

As discussed earlier, the discriminant analysis is undertaken to identify the factorsthat discriminate or predict the OSS status of Indian MFIs. The four significantfactors affecting the sustainability of Indian MFIs and the two component factorsof the OSS ratio are selected as independent variables for the discriminant analysismodel. The pertinence of the two additional component variables included in thisanalysis is discussed below.

4.6.1 Revenue Generation Factor

This factor denotes the means for revenue generation for an MFI, like interest rates,fee incomes and financial margins. The prominence of this factor is discussed byRobinson (1996) and Conning (1999) who observe that only those MFIs whichcharge high and cost-covering interest rates are found to be profitable. Cull et al.(2007) confirm this observation but add to it that if interest rates become exorbitantand exploitative in nature, the MFIs will no longer be profitable as the demandfor microcredit will subside. To pre-empt such an unfortunate situation, whichadversely affects the MFIs and the poor clientele, the Malegam Committee Report(2011) insists a cap on interest rates at 24 % and a financial margin cap at 12 %,for Indian MFIs. Nevertheless, Littlefield and Rosenberg (2004) argue that onlythose MFIs that cover all their expenses by operating at adequate financial marginsare seen to be sustainable. Going by these observations, in this work, financialmargin ratio [Financial margin ratio (FMR) D (Revenue from interest—Financial

Page 75: Sustainability of Indian Microfinance Institutions ||

60 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

expenses)/Average assets] is used to proxy the revenue generation factor. Past worksin similar lines by Crombrugghe et al. (2008) have used the figure of interest rate asa proxy variable for revenue generation factor.

4.6.2 Cost-Efficiency Factor

This factor denotes the efficiency level of MFI operations and is crucial for theOSS of MFIs. Qayyum and Ahmad (2006) confirm this aspect by conducting a DataEnvelopment Analysis study that reports a direct relationship with the efficiency andsustainability of Indian MFIs. Churchill (2000) exhorts MFIs to work towards thegoal of efficiency and cost reduction, by adopting the efficient banking managementpractices. Savita (2007) conducts three case studies on Indian MFIs and opines thatby minimizing the cost per borrower, operating efficiency can be achieved. Goingby this observation this work uses operating cost per borrower [Operating cost perborrower (OCPB) D Operating cost of MFI/Number of borrowers] as a proxy forcost-efficiency. Prior works by Crombrugghe et al. (2008) and Ayayi and Sene(2007) have used operating cost per borrower and total cost ratios to depict thisfactor. For the sake of parsimony of the regression model, this work uses the formervariable alone.

Thus, revenue generation factor and cost-efficiency factor, along with the fourdeterminant factors identified by the regression analysis, are chosen as the probablediscriminator variables for the discriminant analysis model.

4.7 Data and Model for Discriminant Analysis

The sample of 50 MFIs which has disclosed their operational data to MIX for theyears 2005–2009 is used for this analysis too. Discriminant analysis models areformulated and tested on 2007 OSS data. Later these models are validated using2008 and 2009 OSS data, along with their respective lagged predictor variables.

Two discriminant analysis models are framed using the identified six laggedpredictor variables. These two models with 1 year and 2 year time lags are tested on50 Indian MFIs for the year 2007. Among them 43 are operationally self-sustainableand seven are operationally unsustainable. Though this is a small sample size itmeets the sample criteria for discriminant analysis. Discriminant analysis requiresminimum five observations to be used for each independent variable incorporatedin the model, i.e. a minimum ratio of 5:1 has to be kept. In this study there are sixindependent variables and 50 MFIs as observations. This results in a ratio of 8.33:1,which is well within the sample size criteria. In addition to the requirement for theratio of observations to independent variables, discriminant analysis also requiresthat there be a minimum number of observations in the smallest group defined bythe dependent variable. The criterion is that number of observations in the smallest

Page 76: Sustainability of Indian Microfinance Institutions ||

4.8 Multiple Discriminant Analysis Models and Results 61

group must be larger than the number of independent variables. Since the number ofMFIs in the smallest group (seven unsustainable MFIs) is greater than the numberof independent variables used in the model (six variables), this criterion is also met.

The data on these 50 MFIs are used for testing a multiple discriminant analysismodel. Multiple discriminant analysis is a multivariate statistical technique formu-lated by R. A. Fisher in the year 1936. It derives a linear combination of continuousindependent variables that will discriminate best between a priori defined groups;which in this work is a group of operationally sustainable and unsustainable IndianMFIs. This is achieved by maximizing the between-group variance relative to thewithin-group variance. A discriminant function that represents a linear combinationof the independent variables, weighted to maximize the difference between thegroups, is formulated in the analysis. The discriminant function used in this study ismodelled as follows:

D D a C b1X1 C b2X2 C b3X3 C � � � C b6X6 (4.2)

where D D the discriminant score for each MFI in the sample, a is a constant,X1 through X6 represent the six independent variables and b1 through b6 thediscriminant function coefficients or weights for each of the six independentvariable.

The predictive power of this discriminant analysis model was popularized byAltman (1968), who coined the widely used Altman’s Z score model, for predictingthe probability that a firm will go into bankruptcy within 2 years. Akin to this work,this study aims to use discriminant analysis technique to predict the sustainabilitystatus of an Indian MFI.

4.8 Multiple Discriminant Analysis Models and Results

Multiple discriminate analysis was conducted on the sample of 50 Indian MFIsusing SPSS software version 17. A nonmetric category of dependent variablewas created for this purpose, coding operationally self-sustainable MFIs as 1 andunsustainable MFIs as 2 for this purpose. The independent variables, namely,financial margin ratio, portfolio risk ratio, operating cost per borrower, gross loanportfolio and average loan size are measured in metric terms. Credit delivery modelis measured as dummy coded variables by assigning the value of 1 for SHG modeland 0 for Grameen model. The descriptive statistics for the variables in metric termsfor the years 2005–2008 are summarized in Table 4.5.

Table 4.5 shows the financial margin ratio, average loan size per borrower andgross loan portfolio of the MFIs to have increased from the year 2005 to the year2008. Portfolio at risk ratio and operating cost per borrower has shown a decreasein value over the same period.

Two discriminant analysis models are coined based on these dependent andindependent variable combinations.

Page 77: Sustainability of Indian Microfinance Institutions ||

62 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

Tabl

e4.

5D

escr

iptiv

est

atis

tics

for

the

prob

able

disc

rim

inat

orva

riab

les

Dis

crim

inat

ors

Mea

n(2

008)

Stan

dard

devi

atio

n(2

008)

Mea

n(2

007)

Stan

dard

devi

atio

n(2

007)

Mea

n(2

006)

Stan

dard

devi

atio

n(2

006)

Mea

n(2

005)

Stan

dard

devi

atio

n(2

005)

FMR

.084

0.0

4510

.067

4.0

4794

.028

4.0

4638

.045

0.0

8300

PAR

>30

.009

6.0

2016

.011

0.0

2098

.012

4.0

1178

.028

0.0

2350

OC

PB(i

nU

SD)

22.0

800

30.7

7956

25.5

000

38.4

0613

27.4

800

47.9

9245

53.5

800

143.

4463

7A

LSP

B(i

nU

SD)

160.

0200

232.

8899

617

8.84

0025

1.09

185

149.

7400

170.

1028

313

8.38

0017

3.52

538

GL

P(i

nU

SD)

41,7

26,0

0090

,805

,600

26,9

73,0

0049

,512

,100

14,5

13,0

0021

,614

,500

8,84

2,20

016

,628

,900

Page 78: Sustainability of Indian Microfinance Institutions ||

4.8 Multiple Discriminant Analysis Models and Results 63

Model 1

D.2007/ D a C b1X1.2006/ C b2X2.2006/ C b3X3.2006/ C � � � C b6X6.2006/ (4.3)

where D D the discriminant score for each MFI in the sample for the year 2007, a isa constant, X1 through X6 represent the six independent variables for the year 2006and b1 through b6 the discriminant function coefficients or weights for each of thesesix independent variable.

The model attempts to answer the following question:Can the sustainability groups of MFIs for the year 2007 be predicted based on the

six significant variables—financial margin, portfolio at risk ratio, operating cost perborrower, average loan size per borrower, gross loan portfolio and credit model—forthe n�1 period, i.e. for the year 2006?

Model 2

D.2007/ D a C b1X1.2005/ C b2X2.2005/ C b3X3.2005/ C � � � C b6X6.2005/ (4.4)

where D D the discriminant score for each MFI in the sample for the year 2007, a isa constant, X1 through X6 represent the six independent variables for the year 2005and b1 through b6 the discriminant function coefficients or weights for each of thesesix independent variable.

The model attempts to answer the following question:Can the sustainability groups of MFIs for the year 2007 be predicted based on the

six significant variables—financial margin, portfolio at risk ratio, operating cost perborrower, average loan size per borrower, gross loan portfolio and credit model—forthe n�2 period, i.e. for the year 2005?

Before proceeding with the analysis of these two models, the assumptions ofdiscriminant analysis are tested. All independent variables are found to be normallydistributed, expect operating cost per borrower and gross loan portfolio. A logtransformation was done for these two variables, and this ensured compliancewith assumption of multivariate normality. The assumption of linearity betweenvariables is also checked, by plotting a scatter matrix and no non-linear relationshipsare present. Variables did not require further transformation, as there is no traceof multicollinearity among them. The assumption of homogenous dispersion fordependent variable groups is tested using the Box’s M test,3 while performing thediscriminant analysis. A direct method of discriminant analysis is performed forboth the models, where all the six independent variables are simultaneously includedin the analysis to test their significance in discrimination. The Box’s M test failed toreject the null hypothesis of equal dispersion matrices for the dependent variablegroups, confirming equality of group covariance matrices in both the models.

3The null hypothesis of this test is that the dispersion matrices are homogenous. Therefore, if thenull hypothesis is accepted, the assumption of homogenous dispersion matrices across the groupsof the dependent variable is complied with.

Page 79: Sustainability of Indian Microfinance Institutions ||

64 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

Table 4.6 Discriminant analysis estimates

Discriminant analysis Model 1 Model 2

Eigenvalue .549 .629Wilks’ lambda .646* .614*

Squared canonical correlation 35 % (i.e. .595*.595) 39 % (i.e. .621*.621)Original classification accuracy 94 % 96 %Cross-validated classification accuracy 92 % 96 %N (i.e. sample size) 50 50

Note: *Indicates the value is significant at 5 % level

Thus all the assumptions of discriminant analysis are complied with, making theinterpretation of the discriminant analysis outputs valid for both the models.

The analysis resulted in discriminant functions that choose the best combinationof weights that maximizes the differences in discriminant scores (D) between theMFIs in the defined groups of operational self-sustainability. The discriminantanalysis estimates for both the models are summarized in Table 4.6.

The eigenvalues of the discriminant functions are depicted in Table 4.6 as .549and .629. The eigenvalues of the functions depict the ratio of between-group towithin-group sums of squares. As Malhotra (2003) cites a larger eigenvalue to implya superior discriminant function, model 2 appears to have more predictive power.

To check the overall significance of the discriminant functions, the Wilks’lambda statistics is examined. Wilks’ lambda is used to test the null hypothesisthat the means of all of the independent variables are equal across groups of thedependent variable. This hypothesis is rejected in this work, as the chi-squarestatistic corresponding to Wilks’ lambda is statistically significant at 5 % level forboth the models. The values for Wilks’ lambda are depicted in Table 4.6. Sincethese values are significant, it is concluded that there is a relationship between thedependent groups and the independent variables.

Similar to Wilks’ lambda, another indicator that shows the strength of rela-tionship between variables in the model is the canonical correlation coefficient. Itmeasures the association between the discriminant score and the set of independentvariables and its values are .595 and .621, respectively, for the two models. Thesquared canonical correlation depicts the per cent of variation in the dependentvariable, discriminated by the set of independents variables in the discriminantanalysis model. In this study it is 35 % (i.e. 595 * .595) for model 1 and 39 % (i.e..621 * .621) for model 2. This value, though depicts the strength of relationship,does not have any association with the predictive accuracy, which is our ultimatemeasure of the value of the model. The predictive accuracy is denoted by theoriginal and cross-validated classification accuracy values in Table 4.6. The detailedclassification accuracy results for both the models are presented below in Tables 4.7and 4.8, respectively.

The original classification is the classification in which all observations in theanalysis are classified by the function created using all the observations in thesample. Table 4.7 shows that for model 1, out of the 43 MFIs which are operationally

Page 80: Sustainability of Indian Microfinance Institutions ||

4.8 Multiple Discriminant Analysis Models and Results 65

Table 4.7 Classification results for model 1

Classification results

Predicted group membership

Types of classification accuracySustainabilitygroups 2007 Sustainable Unsustainable Total

Original Count Sustainable 43 0 43

Unsustainable 3 4 7

% Sustainable 100:0 :0 100:0

Unsustainable 42:9 57:1 100:0

Cross-validated Count Sustainable 42 1 43

Unsustainable 3 4 7

% Sustainable 97:7 2:3 100:0

Unsustainable 42:9 57:1 100:0

94.0 % of original grouped cases correctly classified92.0 % of cross-validated grouped cases correctly classified

Table 4.8 Classification results for model 2

Classification results

Predicted group membership

Types of classification accuracySustainabilitygroups 2007 Sustainable Unsustainable Total

Original Count Sustainable 43 0 43

Unsustainable 2 5 7

% Sustainable 100:0 :0 100:0

Unsustainable 28:6 71:4 100:0

Cross-validated Count Sustainable 43 0 43

Unsustainable 2 5 7

% Sustainable 100:0 :0 100:0

Unsustainable 28:6 71:4 100:0

96.0 % of original grouped cases correctly classified96.0 % of cross-validated grouped cases correctly classified

self-sustainable, 43 are correctly classified by the model and none of the MFIs areincorrectly classified into the unsustainable group. Out of the 7 MFIs which areoperationally unsustainable, four MFIs are correctly classified and three MFIs areincorrectly classified into the sustainable group. Thus model 1 successfully classifies100 % of the operationally self-sustainable MFIs and 57.1 % of the operationallyunsustainable MFIs, as per the original classification matrix. Over all it correctlyclassifies 47 MFIs out of the total sample size of 50 MFIs, with a classificationaccuracy of 94 %.

The cross-validated classification matrix shows much more realistic classificationaccuracy, as it is computed by sequentially holding out one observation from theanalysis and using the remaining observations to derive the discriminant functionused to classify that observation. Thus it successively classifies all MFIs but one,to develop a discriminant function and then categorizes the MFI that was left out.This process is repeated with each MFI left out in turn. This cross-validation is

Page 81: Sustainability of Indian Microfinance Institutions ||

66 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

considered to produce a more reliable and less biased discriminant function (thoughin most cases it produces a poorer accuracy outcome than the original classification),as in this validation process researcher does not use the case (i.e. the MFI) he/she istrying to predict as part of the categorization process (Malhotra 2003).

For model 1, as per the cross-validated classification matrix shown in Table 4.7,out of the 43 MFIs which are operationally self-sustainable, 42 are correctlyclassified by the model and 1 MFI is incorrectly classified into the unsustainablegroup. Out of the seven MFIs which are operationally unsustainable, four MFIs arecorrectly classified and 3 MFIs are incorrectly classified into the sustainable group.Thus the model successfully classifies 97.7 % of the operationally self-sustainableMFIs and 57.1 % of the operationally unsustainable MFIs, as per the cross-validatedclassification matrix. Over all it correctly classifies 46 MFIs out of the total samplesize of 50 MFIs, with a classification accuracy of 92 %.

Now the interest is to analyse the predictive accuracy of model 2 (as depictedin Table 4.8) to see if this model has a better predictive power than model 1.Table 4.8 shows that for model 2, out of the 43 MFIs which are operationallyself-sustainable, 43 are correctly classified by the model and none of the MFIs areincorrectly classified into the unsustainable group. Out of the seven MFIs whichare operationally unsustainable, five MFIs are correctly classified and two MFIs areincorrectly classified into the sustainable group. Thus model 2 successfully classifies100 % of the operationally self-sustainable MFIs and 71.4 % of the operationallyunsustainable MFIs, as per the original classification matrix. Over all it correctlyclassifies 48 MFIs out of the total sample size of 50 MFIs, with a classificationaccuracy of 96 %.

Now, we check the cross-validated predictive accuracy of model 2. For model2, as per the cross-validated classification matrix, out of the 43 MFIs which areoperationally self-sustainable, 43 are correctly classified by the model and none ofthe MFIs are incorrectly classified into the unsustainable group. Out of the sevenMFIs which are operationally unsustainable, five MFIs are correctly classified andtwo MFIs are incorrectly classified into the sustainable group. Thus the modelsuccessfully classifies 100 % of the operationally self-sustainable MFIs and 71.4 %of the operationally unsustainable MFIs, as per the cross-validated classificationmatrix. Over all it correctly classifies 48 MFIs out of the total sample size of 50MFIs, with a cross-validated classification accuracy of 96 %, which is the same asthat of the model’s original classification accuracy.

Thus the above discussion depicts model 2 to have a superior predictive powerthan model 1, as revealed by the former’s high and consistent predictive accuracyresults. Having compared the model’s predictive power, we are now interested infinding out which among the independent variables used in the discriminant analysismodels (probable discriminators) resulted in this predictive accuracy. Table 4.9presents this by depicting the discriminant coefficients and significance of theindependent variables.

Table 4.9 shows two variables to be significant under model 1—portfolio at riskgreater than 30 days and operating cost per borrower. Their discriminant coefficientsare .214 and .845, respectively. Under model 2, which has a superior predictive

Page 82: Sustainability of Indian Microfinance Institutions ||

4.8 Multiple Discriminant Analysis Models and Results 67

Table 4.9 Discriminantcoefficients Independent

variablesDiscriminantcoefficients (model 1)

Discriminantcoefficients (model 2)

FMR �.314 (.224) �.270 (.874)PAR > 30 .214 (4.882)* .116 (.092)ALSPB �.316(.698) �.533 (.591)CM �.471(2.597) .073 (.031)Log GLP .306(1.650) �.215 (1.195)Log OCPB .845(16.102)* .974 (22.045)*

Note: Figures in parentheses show F values*Indicates that statistically significant difference in dependentvariable groups exists for the independent variables

Table 4.10 Structure matrix

Independent variablesStructure matrix(model 1)

Structure matrix(model 2)

FMR :092 �:170

PAR > 30 :431 :055

ALSPB �:163 �:140

CM �:314 �:032

Log GLP :250 �:199

Log OCPB :782 :855

accuracy, only a single variable is found significant. It is the variable ‘operating costper borrower’, denoting the cost-efficiency factor of an MFI 2 years prior to the dateof estimation. The variable is significant in discriminating the OSS status of IndianMFIs and has a coefficient value of .974.

The signs of discriminant function coefficient are arbitrary and are not indicativeof the direction of relationship. But the coefficient with highest magnitude isregarded the strongest discriminant of the dependent variable groups. As per thiscriterion and by referring to the coefficients of model 2 (which has superior pre-dictive accuracy), operating cost per borrower is the strongest discriminant of OSSgroups among Indian MFIs. This relative importance of operating cost per borroweris also corroborated by the structure matrix, which shows the simple correlationbetween each independent variable and the discriminant function. Operating costper borrower is seen to have the highest magnitude among the structure correlationsfor model 2, as shown below in Table 4.10.

Thus the estimates for model 2 depicts that the cost-efficiency value for an IndianMFI for a given year is a significant discriminant that has the power to predict itsOSS status, 2 years thereafter. Since model 2 has a superior predictive power thanmodel 1, only this factor is regarded as a significant discriminator of an Indian MFI’sOSS status. To validate this finding, the discriminant analysis model 2 is applied tothe years 2008 and 2009. Could we have used this model to predict the OSS statusof Indian MFIs for the years 2008 and 2009, by using data for the years 2006 and2007 respectively (i.e. 2 year’s prior data), then the model’s predictive accuracy canbe verified. For this the classification function coefficient of the model 2 is used,which is presented below in Table 4.11.

Page 83: Sustainability of Indian Microfinance Institutions ||

68 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

Table 4.11 Classificationfunction coefficients Classification function coefficients

Sustainability groups 2007

Sustainable Unsustainable

FMR 2005 8:529 1:277

PAR > 30 2005 34:310 44:804

ALSPB 2005 :002 �:004

CM 2005 1:436 1:756

Log GLP 2005 1:722 1:178

Log OCPB 2005 5:523 10:500

(Constant) �5:148 �13:070

OSS group membership for 2008 predicted using data for 2006 as discriminators

Actual Data for 2008

Predicted Data for 2008

41sustainableMFIs

9unsustainableMFIs

4 out of 9unsustainableMFIs identified asunsustainable

41 out of 41sustainable MFIsidentified assustainable

45/50 MFIscorrectly

classified. i.e.90 per centaccuracy

Fig. 4.3 Prediction accuracy for year 2008

For each of the 50 MFIs in the years 2006 and 2007, SPSS software assignsthe values for independent variables to the classification function coefficients ofthis model to calculate its discriminant functions for the two groups of dependentvariable—sustainable and unsustainable MFI groups. If the MFI has a higher valuefor the group of sustainable MFIs, it is assigned to sustainable MFI group. On thecontrary if the MFI has a higher value for the group of unsustainable MFIs, it willbe assigned to the unsustainable MFI group. This process is repeated for all the 50MFIs for the years 2006 and 2007 using SPSS and the predicted group membershipsfor these MFIs based on model 2 is the received as the output.4

The results obtained using 2006 and 2007 variables as discriminators are shownin Figs. 4.3 and 4.4, respectively.

4Alternatively, the process of classifying new MFIs can also be done using the concept ofMahalanobis distance. A new case (MFI) will have a distance for each of the sustainability groupcentroids (group means) and therefore can be classified as belonging to the group for which itsdistance is smallest. If the discriminant score of a new MFI has a standard deviation of more than1.96 from a group centroid, then it will be regarded to have less than 5 % chance to belong to thatgroup.

Page 84: Sustainability of Indian Microfinance Institutions ||

4.9 Summary 69

OSS group membership for 2009 predicted using data for 2007 as discriminators

Actual Data for 2009

Predicted Data for 2009

42sustainableMFIs

8unsustainableMFIs

4 out of 8unsustainableMFIs identified asunsustainable

42 out of 42sustainable MFIsidentified assustainable

46/50 MFIscorrectly

classified. i.e.92 per centaccuracy

Fig. 4.4 Prediction accuracy for year 2009

Model 2 is able to attain a predictive accuracy of 90 % and 92 % when appliedto the years 2008 and 2009. This validates the model, as it can achieve at least 90 %accuracy in predicting the OSS status of Indian MFIs, with the 2 year’s prior dataon the discriminators.

Thus the findings of this discriminant analysis indicate cost-efficiency factoras the single prominent discriminator of the OSS status of Indian MFIs. Thisfinding fulfils the second objective of this study, which aims to identify the factorsdiscriminating the OSS status of Indian MFIs. Based on the findings of the study,Indian MFI managers are recommended to put in place requisite interventions tooptimize the effect of cost-efficiency factor on the OSS of their MFIs. This canenable Indian MFIs to attain sustainability in their future microfinance operations.

Moreover by using the discriminant model 2, formulated in this paper, and byassigning values for all the six discriminators to the model, Indian microfinancepractitioners can predict the operational self-sustainability status of their MFI, 2years from the date of estimation. Therefore practitioners are recommended to usethis model to predict the OSS status of their MFIs

4.9 Summary

The regression analysis and discriminant analysis conducted in this chapter com-pletes the quantitative phase of this study. The quantitative phase was designedin the three-phased sequential explanatory mixed-methods research framework, toaccomplish the first and second objectives of this research work. Fulfilling thesetwo objectives, the results of the regression and discriminant analysis, identified thefactors affecting and discriminating the OSS status of Indian MFIs. Growth factor,portfolio risk factor, development factor and institutional factor were identifiedas the factors affecting the OSS of Indian MFIs, and cost-efficiency factor was

Page 85: Sustainability of Indian Microfinance Institutions ||

70 4 Quantitative Phase: Identification of Factors Affecting and Discriminating. . .

identified as the factor discriminating its OSS status. Thus with the help of thisquantitative phase the researcher narrows done to five significant factors whichdetermine and discriminate the OSS status of Indian MFIs. As the quantitativephase infers these significant determinants and discriminants to be the factorsthat Indian MFI managers must concentrate on to enhance their MFI’s OSS, aqualitative inquiry is proposed to understand the managerial strategies associatedwith them. The participant MFIs, on whom this qualitative inquiry can be conducted,is identified in the succeeding chapter.

References

Altman EI (1968) Financial ratios, discriminant analysis and the prediction of corporatebankruptcy. J Financ 23(4):189–209

Ashim KK (2010) Sustainability and mission drift in microfinance: empirical studies on mutualexclusion of double bottom lines. Published Doctoral Thesis, Hanken School of Economics,Helsinki

Ayayi A, Sene M (2007) What drives microfinance institution’s financial sustainability. J Dev Area44(1):303–324

Bhatt N, Tang SY (2001) Delivering microfinance in developing countries: controversies and policyperspective. Policy Stud Organ 29(2):319–334

Bogan V (2008) Microfinance institutions: does capital structure matter? www.papers.ssrn.com/sol3/papers.cfm?abstract_id=1144762&rec=1&srcabs=1480844. Retrieved 25 Dec 2010

Christen RP (2001) Commercialization and mission drift. The transformation of microfinancein Latin America. Consultative Group to Assist the Poor (CGAP) Occasional Paper 5,Washington, DC

Churchill C (2000) Banking on customer loyalty. J Microfinanc 2:1–21Coleman AK (2007) The impact of capital structure on the performance of microfinance

institutions. J Risk Financ 8(1):56–71Conning J (1999) Outreach, sustainability and leverage in monitored and peer monitoring lending.

J Dev Econ 60:229–248Crombrugghe AD, Tenikue M, Sureda J (2008) Performance analysis for a sample of microfinance

institutions in India. Ann Pub Coop Econ 79(2):269–299Cull R, Kunt AD, Morduch J (2007) Financial performance and outreach: a global analysis of

leading MicroBank. Econ J 117:107–133D’Espallier B, Guerin I, Mersland R (2011) Women and repayment in microfinance: a global

analysis. World Dev 39(5):758–772Fisher RA (1936) The use of multiple measurements in taxonomic problems. Ann Eugen 7(2):

179–188Ghatak M (2000) Screening by the company you keep: joint liability lending and the peer selection

effect. Econ J 110:601–631Ghatak M, Guinnane TW (1999) The economics of lending with joint liability: theory and practice.

J Dev Econ 60:195–228Gujarati D, Sangeetha S (2007) Basic econometrics. Tata McGraw-Hill, New YorkIntellecap (2010) Indian microfinance crisis of 2010: turf war or a battle of intentions? An

Intellecap white paper. Intellecap, HyderabadLittlefield E, Rosenberg R (2004) Microfinance and the poor: breaking down walls between

microfinance and formal finance. Financ Dev 41(2):38–40Malegam Committee Report (2011) Report of the Reserve Bank of India sub-committee of its

Central Board of Directors to study issues and concerns in the Micro Finance Institutions

Page 86: Sustainability of Indian Microfinance Institutions ||

References 71

(MFI) sector, Reserve Bank of India. http://www.rbi.org.in/scripts/BS_PressReleaseDisplay.aspx?prid=23780. Retrieved 25 Dec 2010

Malhotra KN (2003) Marketing research—an applied orientation. Prentice-Hall, New DelhiMarakkath N, Ramanan TR (2012) What discriminates the operational self-sustainability of

Indian MFIs: a multiple discriminant analysis inquiry. Paper presented at 3rd internationalconference on Institutional and Technological Environment for Microfinance (ITEM3) on CostManagement & Social Performance in Microfinance, New Delhi, 4–7 Jan 2012. Organized byBurgundy School of Business, France (Secured Top Paper Presentation Award)

Nair A (2005) Sustainability of microfinance self help groups in India: would federating help?World Bank Policy Research Working Paper 3516

Navajas S, Schreiner M, Richard M, Claudio G, RodriguezMeza J (2000) Microcredit and thepoorest of the poor: theory and evidence from Bolivia. World Dev 28(2):333–346

Nisha B (2007) Microfinance and microfinance institutions in India: issues and challenges.Network 11(2):1

Olivers-Polanco F (2005) Commercializing microfinance and deepening outreach? Empiricalevidence from Latin America. J Microfinanc 7(2):38–40

Pischke V (1996) Measuring the trade-off between outreach and sustainability of microentrepriselenders. J Int Dev 8:225–239

Qayyum A, Ahmad M (2006) Efficiency and sustainability of micro finance institutions in SouthAsia. MPRA Paper 11674, University Library of Munich, Germany. http://www.saneinetwork.net/pdf/SANEI_VI/SANEI-VI%20PROJECT%207%20Efficiency%20and%20Sustainability%20of%20Micro%20Finance%20Institutions%20in%20South%20Asia.pdf. Retrieved 25 Dec2010

Robinson SM (1996) Addressing some key questions on finance and poverty. J Int Dev 8:153–163Savitha S (2007) Transaction cost in group microcredit in India. Case studies of three micro finance

institutions. Working paper series. The Institute of Financial Management and Research, Centreof Microfinance, Chennai, India. http://www.ifmr.ac.in/cmf/publications/wp/2006/13_shankar-caseStudyMFIs.pdf. Retrieved 6 May 2010

Venkatraman S, RajSekhar T (2008) For India’s microfinance institutions, governance is the keyto sustained and scalable growth.www.standardandpoors.com/ratingsdirect. Retrieved 25 Dec2010

Page 87: Sustainability of Indian Microfinance Institutions ||

Chapter 5Intermediate Participant Selection Phase:Assessment of Efficiency and Sustainability

5.1 Preface

The preceding chapter undertook a quantitative analysis to identify the determinantsand discriminant of the OSS status of Indian MFIs. This phase conducts a perfor-mance analysis on Indian MFIs, to identify the operationally efficient MFIs, whichremain sustainable by charging a reasonable interest rate. The identified MFIs willthen be regarded as participants for the ensuing qualitative analysis. This chaptertherefore is dedicated to discuss the intermediate participant selection phase of thestudy.

5.2 Structure of the Intermediate Participant Selection Phase

The structure of the processes involved in the intermediate phase is portrayed inFig. 5.1. This phase intends to fulfil the third objective of the study—identificationof the efficient and sustainable Indian MFIs. In order to formulate a model forassessing the efficiency of MFIs, firstly a literature review is undertaken. Based onthis literature survey undertaken, the input–output specification for the DEA modelis decided. Data is sourced on these input–output variables from 50 Indian MFIs andthe DEA model is tested on this data to obtain the relative efficiency scores for thesample MFIs. The results of this efficiency analysis identify the relatively efficientMFIs operating in the industry. Thereafter a sustainability assessment model isformulated to assess the sustainability of the efficient MFIs. Thus as a result ofthis sustainability assessment, efficient MFIs which remain sustainable by levying

This chapter has materials published by the author in the article: ‘Assessing the Efficient andSustainable Performance of Indian Microfinance Institutions’. Marakkath N, Ramanan RT (2012)Cost Manage. Thomson Reuters/RIA, 26 (5): 1–14

N. Marakkath, Sustainability of Indian Microfinance Institutions:A Mixed Methods Approach, India Studies in Business and Economics,DOI 10.1007/978-81-322-1629-2__5, © Springer India 2014

73

Page 88: Sustainability of Indian Microfinance Institutions ||

74 5 Intermediate Participant Selection Phase: Assessment of Efficiency. . .

Intermediate Phase

Data Envelopment Analysis Model

Sustainability Diamond Model

Efficient Peer Indian MFIs

Efficient & Sustainable Peer Indian MFIs

Objective 3: To Identify Efficient and Sustainable Indian MFIs, Which Can Act asReference Group

Literature Review on MFI Efficiency Analysis

Data Collection & Analysis

Assess influence of institution-specific factors onefficient and sustainable Indian MFIs

Fig. 5.1 Structure of the intermediate phase

a reasonable interest rate from the poor are found out. Subsequently, regression anddiscriminant analysis are undertaken to assess whether institutional specific factorshave an influence on the efficient and sustainable status of the identified MFIs.

The processes sketched in Fig. 5.1 are elucidated in the ensuing sections of thischapter.

5.3 Literature Review on Microfinance InstitutionEfficiency Analysis

Authors like Farrington (2000) and Lafourcade et al. (2005) have used ratioanalysis technique for MFI efficiency measurement. Stochastic frontier analysis, aparametric method, is used by authors like Hassan and Tufte (2001) and Desrochersand Lamberte (2003) for efficiency analysis. But both ratio analysis and stochasticfrontier analysis techniques have limitations in using multiple inputs and multipleoutputs for estimating the joint efficiency of MFIs. This can be effectively done byDEA, a non-parametric method that does not impose a priori functional form forproduction technology. Despite this advantage, DEA is used only in a handful ofstudies to examine the efficiency of MFIs. Some attempts made across the world inthis direction are discussed below.

Page 89: Sustainability of Indian Microfinance Institutions ||

5.4 Data Envelopment Analysis Methodology 75

Nghiem et al. (2006) conducted a DEA study on 46 MFIs in Vietnam, usinglabour cost and nonlabour costs as inputs and number of savers, number ofborrowers and number of groups as outputs. The 46 MFIs were found technicallyefficient, with an average technical efficiency score of 80 %. Guitierrez-Nieto et al.(2007) conducted a similar DEA work on 30 Latin American MFIs. Number ofcredit officers and operating expenses were used as inputs and number of loansoutstanding, gross loan portfolio and interest and fee income were used as outputs.Qayyum and Ahmad (2006) also contributed in this line, by using an extendedsample of 85 MFIs from South Asia for DEA analysis. Fifteen Pakistani, 25 Indianand 45 Bangladeshi MFIs were assessed using a DEA model, which had creditofficers and cost per borrower as inputs and loans disbursed by MFI as output. Thestudy attributes inefficiencies in the three South Asian countries to be technical innature than scale inefficiencies and calls for more managerial and technologicalimprovements. Another DEA work was conducted on 20 Malaysian MFIs by Sufianin the year 2006. Total deposits and fixed assets were used as inputs and total loansand other income were used as outputs in the DEA model. The study observed only28.75 % of all Malaysian MFIs to be efficient and more profitable. In the year 2008,35 MFIs in Mediterranean region were assessed using a DEA model in a work byBassem. Number of personnel and total assets were used as inputs and number ofwomen borrowers and return on assets were used as outputs in the model. The studyfound eight MFIs in the Mediterranean region to be technically efficient. Recently,in the year 2009, Haq, Skully and Pathan conducted a DEA study on 39 MFIs acrossAfrica, Asia and Latin America. Labour, cost per borrower and cost per saver wereused as inputs and savers per staff member and borrowers per staff member wereused as outputs. The study concluded by commenting that in the long run, bankMFIs will outperform NGO MFIs, as they have more access to local capital market.

Out of the above-discussed works, only the work done in India, by Qayyumand Ahmad (2006), follows up the DEA efficiency analysis with a sustainabilityassessment, using scale parameter. But the work does not undertake a benchmarkinganalysis. None of the other works have adopted a benchmarking exercise to identifya set of efficient peers from the sample MFIs and have thereafter attempted toascertain its sustainability levels. In this respect the efficiency analysis undertaken inthis study is the first of its kind in microfinance literature. Apart from this comparedwith existing literature, this study is more comprehensive, as the efficiency analysismodel and sustainability assessment model framed in this study captures both thesocial and financial performance of Indian MFIs. Before discussing about thesemodels, an introduction to the DEA methodology is given in the next section.

5.4 Data Envelopment Analysis Methodology

DEA is a linear programming methodology, popularized by Charnes et al. (1978),by building on the efficiency ideas put forth by Farrell (1957). It is a data orientedapproach for evaluating the performance of a set of entities which convert multiple

Page 90: Sustainability of Indian Microfinance Institutions ||

76 5 Intermediate Participant Selection Phase: Assessment of Efficiency. . .

Fig. 5.2 Concept of dataenvelopment analysis(Source: Coelli 1996;Ramanathan 2003)

inputs into multiple outputs. DEA evaluates the efficiency of each entity throughan objective weight system that the model calculates from the data, without relyingon the subjective opinions of researchers. The model assigns that set of weightsto an entity that maximizes its efficiency score. This method is widely acceptedamong strategic, policy and operational circles, particularly in the service andnonprofit sectors (Siems and Barr 1998). Its wide acceptance is due to its abilityto estimate efficiency scores for complex multi-input or multi-output firms, wherethe underlying production process is not well understood. Since this work intendsto assess the relative efficiency scores of Indian MFIs, whose production processcannot be analytically represented, the DEA method is found most suitable for thispurpose.

DEA incorporates input–output data of entities without requiring a priori spec-ification of functional forms and empirically constructs a production frontier fromthis data using piecewise linear programming methods. The concept of DEA can beexplained using Fig. 5.2.

Figure 5.2 shows a graphical representation of a DEA analysis undertaken usingtwo inputs (x1, x2) and one output (y). In Fig. 5.2, the axes represent ratios of inputto output. The entities that are more efficient use lower levels of input per unit ofoutput and hence are lying closer to the origin. All these operationally efficiententities together constitute the efficient frontier. Efficiency of all others entities isthen measured in terms of how far they are from the efficient frontier. Entities onthe frontier are considered 100 % efficient and those that do not lie on the efficiencyfrontier are regarded relatively less efficient. The efficient entities are regarded asbest performing entities and are assigned an efficiency score of unity or 100 %.

The efficiency scores of other entities vary between 0 and unity or 100 % relativeto the best performance efficiency scores in the sample (Ramanathan 2003). Themethod is called ‘Data Envelopment Analysis’ as the efficiency frontier envelopsthe available data and forms the basis of the efficiency measurement undertaken inthe analysis. The region wherein all the data points are enclosed by the frontierline is called the ‘production possibility set’. An inefficient entity can be made

Page 91: Sustainability of Indian Microfinance Institutions ||

5.4 Data Envelopment Analysis Methodology 77

efficient by projecting it onto a point on the efficient frontier. In Fig. 5.2, considerthe inefficient entity D. To measure its efficiency, a line is projected to the efficiencyfrontier. The projected line is represented in the figure as ‘line OD’ which crossesthe frontier line at P. Then the relative efficiency of entity D is the ratio of distanceOP to the distance OD. This means that the inefficiency of D is to be evaluatedby a combination of A and B because the point P is on the line connecting thesetwo points. A and B are called the ‘reference set’ for D. The entities constitutingthe reference set are called ‘efficient peers’, whose practices the inefficient entityshould emulate in order to enhance its efficiency. So in this example, by calculatingthe extent of inefficiencies present in the current operations of entity D and byidentifying the peers whose practices D should emulate, entity D can be guidedon its path to attain efficiency.

The DEA method can be used under both constant returns to scale and variablereturns to scale assumptions. Constant returns to scale (CRS) reflects the factthat output will change by the same proportion as inputs are changed, whereasvariable returns to scale (VRS) reflects the fact that changing all inputs by thesame proportion in a production technology may exhibit increasing, constant anddecreasing returns to scale. Usage of these two assumptions have led to twopopular models in DEA—the constant returns to scale model, called Charnes,Cooper and Rhodes model (CCR model), and the variable returns to scale model,called Bankers, Charners and Cooper Model (BCC model) (Charnes et al. 1978;Bankers et al. 1984). Both these models of DEA are used in this study. Themathematical formulation of constant returns to scale DEA model and variablereturns to scale DEA model under input orientation is discussed in the appendix(refer to Appendix 2).

These two models are used in this study under input-oriented version. The inputorientation version is used in this work, as it depicts the minimization of inputspossible to produce specified levels of outputs. The output orientation version,which depicts the maximization of outputs, with specified levels of inputs, is notfound appropriate for microfinance industry, which has a social goal of povertyalleviation. Maximizing outputs like interest rates, loan portfolio and financialmargins, without any discretion may amount to client exploitation. Moreover inDEA methodology, if the managers have more control over inputs than outputs,then input orientation is regarded more appropriate.

Thus by using the constant returns to scale DEA model and variable returns toscale DEA model under input orientation, the study obtains both technical efficiencyand scale efficiency scores for the sample MFIs. CRS efficiency score from CCRmodel represents technical efficiency (TE) of an MFI. TE is the overall (gross)efficiency of a firm which comprises of both pure technical efficiency (PTE) andscale efficiency (SE) aggregated into one composite score. TE D PTE � SE. TEin DEA reflects the ability of a entity to obtain maximal output from a given setof inputs relative to the best practice in the sample of entities. But under inputminimization it would denote how a given level of output can be obtained usingminimum inputs. Scale efficiency is the efficiency of the entity when its size ofoperation is assumed to be optimal. Scale efficiency is calculated by dividing CRS

Page 92: Sustainability of Indian Microfinance Institutions ||

78 5 Intermediate Participant Selection Phase: Assessment of Efficiency. . .

efficiency by VRS efficiency. Thus SE D CRS Efficiency/VRS efficiency. VRSefficiency is the efficiency score from BCC model. It measures only pure technicalefficiency (PTE). PTE takes into account the variation in technical efficiency withrespect to scale of operation (Coelli 1996).

5.5 Sample Data and Specification of Inputs and Outputsfor the Data Envelopment Analysis Model

The secondary data on 50 Indian MFIs is collected for the year 2009 for thepurpose of the study. The data is sourced from MIX database. Since the inputsand outputs specification for the DEA model has to be in conformity with thisapproach chosen for doing a DEA, first the DEA approaches applicable to financialinstitutions are identified. Berger and Humphrey (1997) suggest two approaches—production approach and financial intermediation approach—to be commonly usedfor efficiency analysis among financial institutions. The approach chosen for effi-ciency analysis in these financial institutions depends upon what these institutionsactually do.

Going by this logic, the authors try to portray what MFIs do under each of theseapproaches. In a pure production approach, an MFI is assumed to be producersof loans and deposits. That is, in this approach loans and deposits are treated asoutputs, with labour and other capital resources forming the inputs (Soteriou andZenios 1999; Vassiloglou and Giokas 1990). But in a pure financial intermediaryapproach, an MFI is assumed to be an intermediary who makes profits by matchingdepositors and borrowers in a financial market. In this approach, deposits aretreated as inputs, with a surplus generation as output (Berger and Mester 1997;Athanassoupoulos 1997).

Thus it is noted that deposits are treated in two different manner under these twoapproaches. This is not a concern in this study as only a limited number of IndianMFIs (only licensed non-banking financial companies, which have investment-gradecredit rating) are permitted to raise deposits in India. Thus as deposits do notconstitute a homogeneous variable across all MFIs, it does not feature as an input oroutput for this study.1

Since deposits do not constitute a variable for this study, neither a pure productionapproach nor a financial intermediation approach could be adopted. Cingi andTarim (2000) and Guitierrez-Nieto et al. (2007) advocate the usage of a mixture ofboth these approaches for efficiency analysis of financial institutions. So similar toGuitierrez-Nieto et al. (2007), a mixture of both these approaches is adopted in thisstudy. The DEA model proposed in this study views MFIs as financial institutionsbound to keep its dual goals—both social and financial (Woller et al. 1999; Schreiner2002; Guitierrez-Nieto et al. 2008). Thus social and financial goals of an MFI formsthe outputs for the DEA model used in this study.

1DEA requires homogeneous data for all entities/firms under study.

Page 93: Sustainability of Indian Microfinance Institutions ||

5.5 Sample Data and Specification of Inputs and Outputs. . . 79

OUTPUTSINPUTS

NUMBEROF WOMEN

BORROWERSTOTALASSETS SOCIAL

OUTPUTSNUMBEROF POOR

BORROWERSMFI

OPERATINGCOST PER

BORROWER

GROSSLOAN PORTFOLIO FINANCIAL

OUTPUTS

INTERESTAND FEES INCOME

NUMBEROF CREDITOFFICERS

Fig. 5.3 Data envelopment analysis model (Source: Marakkath and Ramanan 2012)

The social goal is denoted by depth of outreach, i.e. the extent to whichmicrofinance reaches the poor. Depth of outreach can be captured by poverty leveland gender of the clients (Christen 2001; Navajas et al. 2000; Bhatt and Tang 2001).The assumptions are that the greater the number of poor clientele and womenclientele served by microfinance, the deeper is the outreach. Both these variablesare included as outputs in the DEA model, as per production approach.

The financial goal on the other hand is denoted by the MFI’s ability to generatea surplus on its growing loan portfolio (Otero 2000; Robinson 2001). These arecaptured by the gross loan portfolio of an MFI and the interest and fee incomecharged by them.2 Gross loan portfolio is included as an output in the model as perproduction approach and interest and fee income is included as per intermediationapproach.

The input specification in this model has three variables—total assets, numberof credit officers and operating cost per borrower. The former two variables areincluded as per production approach and the latter as per intermediation approach.This is so as these variables serve as inputs for an MFI’s operations, as per theserespective DEA approaches. These variables also represent the inputs for threecore performance indicators of an MFI—return on assets, credit officer productivityand operational self-sustainability, respectively (Micro Rate and Inter-AmericanDevelopment Bank 2003). The DEA model formulated is given below in Fig. 5.3.

Table 5.1 gives the definition of the input and output variables used in the DEAmodel.

2Charging cost-covering interest rates is a means of furthering the financial goal of an MFI. But asdiscussed earlier, if the MFI passes on its operational inefficiencies to the clients, in the form ofhiked interest rates, then it can prove counterproductive for its long-run sustainability.

Page 94: Sustainability of Indian Microfinance Institutions ||

80 5 Intermediate Participant Selection Phase: Assessment of Efficiency. . .

Table 5.1 Definitions for inputs and outputs (Source: Marakkath and Ramanan 2012)

Inputs Definitions

Total assets Net assets of the MFIOperating cost per borrower Operating cost of MFI divided by total number of borrowers of

the MFINumber of credit officers Total number of full time staff of the MFI who work on field for

loan disbursement and loan collection, regardless ofemployment status (contractual or regular)

Output DefinitionsNumber of women borrowers Total number of women clientele of the MFINumber of poor borrowers Total number of poor borrowers of the MFI, calculated as per

the average loan size formulaeGross loan portfolio Total loan outstanding for the MFIInterest and fee income Total operating income for the MFI from interest and fee for

services on loan and investments

The correlations between the input–output variables were checked and since theywere correlated it was considered appropriate for usage in the DEA model. Therelative efficiency scores of Indian MFIs were assessed by testing this DEA modelon a sample of 50 Indian MFIs. In DEA, the number of entities is expected to belarger than the product of number of inputs and outputs in order to discriminateeffectively between efficient and inefficient entities (Avkiran 2001; Darrat et al.2002). As the sample size of 50 MFIs is larger than the product of number of inputsand outputs used in this study, the sample size requirements were complied with.

The relative efficiency score for MFIs were computed using Data EnvelopmentAnalysis Program (DEAP), by comparing a given MFI to a pool of well-performingMFIs that serve as a benchmark for the MFI under evaluation. Data for all thevariables in the model are sourced from the financial statements of the MFIs, exceptthe number of poor borrowers which is not readily available. The data for numberof poor borrowers is calculated from the value of average loan size per capita grossnational income (GNI), using the premise stated by Guitierrez-Nieto et al. (2008).The premise is as follows: ‘Given any two MFIs with identical inputs, the one thatmakes many small loans (small relative to the country’s per capita GNI) will bemore socially efficient that the one that makes larger loans’. Based on this premisethe equation used for deriving the poor borrowers figure is as follows:

pi D 1 � Ki � Minimum .K/

Maximum .K/ � Minimum .K/

Pi D pi�Bi

where Ki D Average Loan Size for the ith MFIPer Capita Gross National Income

pi D proportion of poor borrowers for the ith MFI, 0 < pi < 1Pi D number of poor borrowers for the ith MFIBi D total number of borrowers for the ith MFI

Page 95: Sustainability of Indian Microfinance Institutions ||

5.6 Empirical Analysis and Results 81

Table 5.2 Descriptivestatistics for inputsand outputs

Inputs and outputs Mean Std. deviation

Total assets 88,723,996 184634867.4Operating cost per borrower 18 16.71Number of credit officers 1,665 3483.51Number of women borrowers 439,647 1017817.87Number of poor borrowers 422,591 945013.0539Gross loan portfolio 71,010,676 181225055.5Interest and fee income 15,746,799 38256328.28

Table 5.2 shows the descriptive statistics associated with the input–outputvariables.

The descriptive statistics associated with these variables depict that there isheterogeneity in the values of the inputs and outputs. But in DEA as long as theentities are into a homogenous function, with homogeneous variables, their relativeefficiency can be assessed.

5.6 Empirical Analysis and Results

The empirical analysis done in the intermediate phase is categorized into threeheads: (a) efficiency analysis, (b) benchmarking exercise and (c) sustainabilityassessment.

5.6.1 Efficiency Analysis

Efficiency analysis is undertaken using DEA technique. DEA is performed usinginput orientation version under both CCR model and BCC model.

The CCR model assumes constant returns to scale relationship between inputsand outputs and calculates the overall efficiency for each unit, where both puretechnical efficiency and scale efficiency are aggregated into one value. But theBCC model which assumes variable returns to scale, calculates the pure technicalefficiency alone. The efficiency scores derived from both these models, under inputorientation method is presented in Table 5.3.

From the Table 5.3 the DEA efficiency scores of the MFIs appearing efficientacross both CCR and BCC models, under both input orientation method is noted.Asmita, Bandhan, Kotalipura, Mahasemam, MMFL, Pustikar, Sanghamithra, Sar-vodaya Nano Finance, Share MACTS, SKDRDP, SKS and Spandana were found tobe efficient under both the models as they depicted an efficiency score of 100 %, i.e.a value of 1. Thus 12 MFIs appeared efficient across both the models. The rest ofthe MFIs which have an efficiency score of less than 1 are regarded less efficient.

Page 96: Sustainability of Indian Microfinance Institutions ||

82 5 Intermediate Participant Selection Phase: Assessment of Efficiency. . .

Table 5.3 Data Envelopment Analysis efficiency scores

Input orientation

Serial number MFIs CRSTEa VRSTEb SCALEc irs/drs/crsd

(Column 1) (Column 2) (Column 3) (Column 4) (Column 5) (Column 6)

1 Adhikar 0.783 0.807 0.97 irs2 Asmita 1 1 1 crs3 Asomi 0.794 0.837 0.949 irs4 AWS 0.711 0.957 0.743 irs5 Bandhan 1 1 1 crs6 BASIX 0.615 0.621 0.989 drs7 BFL 0.838 0.84 0.997 irs8 BISWA 0.688 0.691 0.996 irs9 BSS 0.817 0.836 0.978 drs10 Casphor MC 0.916 0.974 0.941 drs11 Cresa 0.746 0.796 0.938 irs12 ESAF 0.76 0.774 0.982 drs13 GSFL 0.903 0.904 0.999 irs14 GSGSK 0.824 0.875 0.942 irs15 GU 0.805 0.826 0.975 irs16 GV 0.9 0.951 0.947 drs17 HIH 0.415 0.415 0.998 irs18 INDUR MACS 0.684 0.763 0.897 irs19 Janalakshmi

FinancialServicesPvt. Ltd.

0.611 0.634 0.964 irs

20 Janodaya 0.954 1 0.954 irs21 JFSL 0.907 0.916 0.99 drs22 Kotalipura 1 1 1 crs23 KSBLAB 0.65 0.673 0.966 irs24 Mahashakti 0.86 0.969 0.887 irs25 Mahesmam 1 1 1 crs26 MMFL 1 1 1 crs27 NBJK 0.801 1 0.801 irs28 NEED 0.805 0.857 0.938 irs29 Nidhan 0.291 1 0.291 irs30 Pusthikar 1 1 1 crs31 PWMACS 0.669 0.792 0.845 irs32 RASS 0.913 0.925 0.987 irs33 RGVN 0.844 0.846 0.998 irs34 Sadhana 0.89 0.945 0.942 irs35 Sanghamithra 1 1 1 crs36 Sarvodaya

NanoFinance

1 1 1 crs

37 SCNL 0.647 0.651 0.994 irs38 SEWA Bank 0.594 0.601 0.988 irs

(continued)

Page 97: Sustainability of Indian Microfinance Institutions ||

5.6 Empirical Analysis and Results 83

Table 5.3 (continued)

Input orientation

Serial number MFIs CRSTEa VRSTEb SCALEc irs/drs/crsd

(Column 1) (Column 2) (Column 3) (Column 4) (Column 5) (Column 6)

39 Share MACTS 1 1 1 crs40 SKRDP 1 1 1 crs41 SKS 1 1 1 crs42 Sonata 0.662 0.665 0.995 irs43 Spandana 1 1 1 crs44 Star 0.918 1 0.918 irs45 SU 0.958 0.978 0.979 irs46 Swadhaar 0.606 0.656 0.924 irs47 SWAWS 0.998 1 0.998 irs48 Ujjivan 0.851 0.923 0.922 drs49 VFS 0.774 0.844 0.917 drs50 WSE 0.98 1 0.98 irs

Mean 0.828 0.875 0.948

Note: The 50 MFIs are those which have disclosed their financial data to MIX Market database inthe year 2009. The ticker symbol of these MFIs as given in MIX Market database is used to denotethem. More details on the identity of these MFIs can be obtained at MIX Market website. http://www.mixmarket.org/mfi/country/IndiaaCRSTE denotes the constant returns to scale technical efficiency. It is the gross efficiency scoreproduced by CCR model under CRS assumption with input orientation. It comprises of scaleefficiency and technical efficiency aggregated into onebVRSTE denotes the variable returns to scale technical efficiency. It is the pure technical efficiencyscore calculated by BCC model under VRS assumption and input orientation. It takes into accountthe variation in technical efficiency with respect to scale of operation. A unit is said to betechnically efficient if it minimizes input per unit of output producedcSCALE denotes the efficiency of a unit calculated when its size of operation is optimal underinput orientation. Scale efficiency is calculated by dividing aggregate efficiency (from the CCRmodel) by technical efficiency (from the BCC model)dirs denotes MFIs with increasing returns to scale, drs denotes MFIs with decreasing returns toscale, and crs denotes MFIs with constant returns to scale

The average input-oriented technical efficiency, pure technical efficiency andscale efficiency are found to be 82.8, 87.5 and 94.8, respectively, as shown in thelast row of column 3, 4 and 5 in Table 5.3. Thus by taking the deviations of theaverage pure technical efficiency score (i.e. 87.5 as given in last row of column4 of Table 5.3) from cent per cent efficiency value (i.e. 100 %), it is concludedthat 12.5 % of inputs can be decreased without affecting the existing output levelsof Indian MFIs. Similarly by taking the deviations of the average scale efficiencyscore (i.e. 94.8 as given in last row of column 5 of Table 5.3) from cent per centefficiency value, it is concluded that there is a different of 5.2 % between actualscale of operations and optimal scale of operations for the sample.

To determine whether MFIs are operating at increasing, decreasing or constantreturns to scale, an additional DEA problem with non-increasing returns to scale(NIRS) was run.

Page 98: Sustainability of Indian Microfinance Institutions ||

84 5 Intermediate Participant Selection Phase: Assessment of Efficiency. . .

If NIRS Technical Efficiency D VRS Technical Efficiency; then MFI is categorizedas operating under decreasing returns to scale .drs/

If NIRS Technical Efficiency ¤ VRS Technical Efficiency; then MFI is categorizedas operating under increasing returns to scale .irs/

If CRS Technical Efficiency D VRS Technical Efficiency; then MFI is categorizedas operating under constant returns to scale .crs/

Based on the above criteria, the DEA analysis identifies the returns to scaleof the MFIs. Table 5.3 shows 46 % of MFIs under input-oriented method to beexperiencing increasing returns to scale. Twenty-eight per cent are shown to beexperiencing decreasing returns to scale and 26 % constant returns to scale. TheMFIs operating under decreasing returns to scale are recommended to curtail theirinput expenditure as it is not contributing positively to its outputs. The MFIsoperating at constant returns to scale are the efficient MFIs, which operate at themost productive scale size (MPSS). MPSS refers to the point on the efficient frontierat which maximum average productivity is achieved for a given input–output mix(Banker et al. 1984). MPSS is the ideal scale size at which the MFI enjoys maximumpossible economy of scale. Beyond this point, the decreasing scale of returnsoperates. Thus the less efficient MFIs operating on decreasing and increasing returnsto scale are recommended to refer to the MPSS of their efficient peers to optimizetheir scale size. The efficient peers are identified in the next section.

5.6.2 Benchmarking Exercise

The DEA analysis also does a benchmarking to identify the peers for each of theless efficient MFIs. DEA identifies for each of the less efficient MFIs a referenceor peer group, which includes those MFIs that are efficient if evaluated with theoptimal system of weights of an inefficient MFI. The peer group, made up of MFIswhich are characterized by operating methods similar to the less efficient MFI beingexamined, is a realistic term of comparison which the MFI should aim to imitatein order to improve its performance. Thus peers are those MFIs which the lesserefficient MFIs should ideally emulate to enhance its efficiency.

In the DEA conducted, the frontier against which the efficiency of all MFIsis measured is defined by those MFIs in the group with an efficiency score of100 %, i.e. a value equal to 1. The MFIs operating on the efficient frontier areconsidered to define the best practices in the microfinance industry. Thus theycan be regarded as reference groups for the rest of the lesser efficient ones. Foreach of the less efficient MFIs, the DEA model has identified efficient MFIs thatcould be used as comparators. These efficient peers are comparable to the lessefficient MFIs based on the input–output specifications of the DEA model used inthe study. The less efficient MFIs are expected to learn from their efficient peers byunderstanding their practices. As per the DEA analysis, nine among the identified 12

Page 99: Sustainability of Indian Microfinance Institutions ||

5.6 Empirical Analysis and Results 85

efficient MFIs appear as peers to other lesser efficient MFIs. Three efficient MFIs—Asmita, MMFL and Share MACTS, are not regarded as peers because as per theDEA models used in this study, the weights assigned to these three MFIs are notcomparable to other lesser efficient MFIs in the sample. The nine efficient peerMFIs are Bandhan, Kotalipura, Mahasemam, Pustikar, Sanghamithra, SarvodayaNano Finance, SKDRDP, SKS and Spandana. The three efficient MFIs which arenot peers to other MFIs are not taken for further analysis in this study, because theaim of this work is to identify a set of efficient and sustainable peer MFIs, whosepractices can be referred to or emulated by other lesser efficient MFIs.

5.6.3 Sustainability Assessment

The analysis also does a sustainability assessment on the identified nine efficientpeer MFIs. The intention is to identify a set of Indian MFIs, which are efficientpeers and at the same time sustainable in its operations. Sustainability assessment isdone in this study by framing a model called ‘Sustainability Diamond’, which takescare of the four dimensions of an MFI’s sustainability—financial, risk, growth andsocial dimensions.

Financial dimension is captured by the OSS ratio of the MFI. As per the CGAPand MIX standards, an OSS ratio of 100 % and above denotes the operationalsustainability of an MFI. Therefore in this study MFIs meeting this standard isregarded to be moving towards the goal of financial sustainability.

Risk is captured by an MFI’s portfolio at risk greater than 30 days ratio. Portfolioat risk greater than 30 days ratio is considered to have a negative relationshipwith sustainability of an MFI. As per CGAP and Micro Rate and Inter-AmericanDevelopment Bank, a portfolio at risk greater than 30 days ratio, exceeding 10 %, isa concern for an MFI’s sustainability, because unlike commercial loans, MFI loansare not backed by collaterals. So, in this study MFIs with portfolio at risk greaterthan 30 days less than 10 % are considered to be moving towards sustainability, witha less risky portfolio.

Growth dimension is captured by the gross loan portfolio to assets ratio of theMFI. This denotes the ratio of loans outstanding to the total assets of an MFI.It is calculated as follows: gross loan portfolio of an MFI (i.e. all outstandingprincipals due for all outstanding client loans, which includes current, delinquentand renegotiated loans, but not loans that have been written off and interestreceivable)/total assets of an MFI. Total asset base is used to control for the sizeof the MFI. A ratio of 50 % denotes that serious efforts are being done by the MFIto increase scalability of its microfinance operations, i.e. to reach out to the poor(Kalim and Salahuddhin 2011). Therefore in this study an MFI with 50 % or abovegross loan portfolio to total assets ratio is considered to be treading on the path tosustainability, by achieving a reasonable balanced growth relative to its asset base.

Social dimension is captured by the yield on gross loan portfolio of an MFI.Yield on gross loan portfolio is used as a proxy for the interest rate charged byan MFI. Lower yield denotes that an MFI charges lower interest rate from the

Page 100: Sustainability of Indian Microfinance Institutions ||

86 5 Intermediate Participant Selection Phase: Assessment of Efficiency. . .

0

1

2

3

4

5

Operational Self-Sustainability

Ratio

Gross LoanPortfolio to Total

Assets Ratio

Yield on GrossLoan Portfolio

Portfolio at Risk >30 days Sustainable MFI

Fig. 5.4 Sustainability diamond model (Source: Marakkath and Ramanan 2012)

poor clientele. For MFIs with OSS greater than 100 %, the one with lower yieldis considered to be more socially sustainable, as it levies lower interest rate fromthe poor clients. In India MFIs used to levy different interest rates until a cap oninterest rate at 26 % was imposed by the Malegam Committee—a special committeeappointed by RBI in the year 2011 to address the post-microfinance crisis issues inthe nation, regarding coercive collection practices, usurious interest rates and over-indebtedness. The committee regards 26 % as a reasonable interest rate that can belevied by MFIs in Indian contexts. Therefore in this study an MFI with yield equalto or less than 26 % is considered to be treading on the path to sustainability, bycharging a reasonable interest rate that does not exploit the poor.

Integrating all these four dimensions, a sustainable MFI is portrayed as followsin Fig. 5.4.

A sustainable MFI is depicted in Fig. 5.4 as a diamond that scores a value of5 on all four parameters discussed above, on a scale of 1–5. As per this model anMFI can score a value of 5 on all these four parameters, if its OSS ratio is 100 % orabove, its portfolio at risk greater than 30 days ratio is 10 % or below, its gross loanportfolio to total assets ratio is 50 % or above and yield on gross loan portfolio is26 % or below, for the assessment year.

The sustainability parameters and scales for the nine efficient peer MFIs identi-fied in this study are depicted in Table 5.5.

As per the scaling, out of the nine efficient peer MFIs, seven are sustainable witha value equal to five for all the four dimensions of an MFI’s sustainability. The sevenefficient and sustainable peer MFIs are Bandhan, Pustikar, Sanghamithra, SarvodayaNano Finance, SKDRDP, SKS and Spandana.

The scaling pattern is discussed in detail in Table 5.4.Though Kotalipura and Mahasemam are efficient peer MFIs, they have a low

social dimension and are therefore not included in this list. Their yield is not within

Page 101: Sustainability of Indian Microfinance Institutions ||

5.6 Empirical Analysis and Results 87

Table 5.4 Scaling pattern for sustainability diamond (Source: Marakkath and Ramanan 2012)

OSS ratioPortfolio ratio at risk>30 days

Gross loan portfolioto total assets ratio

Yield on gross loanportfolio

100 % or above D 5 10 % or below D 5 50 % or above D 5 26 % or below D 5Others D Value*5/100 Others D (10/Value)*5 Others D Value*5/50 Others D (26/Value)*5

Table 5.5 Sustainability parameters and scales for the nine efficient peer microfinance institu-tions (Source: Marakkath and Ramanan 2012)

Serialnumber MFI

OSS ratio ason 1.4.2009

Portfolio atrisk >30 daysratio as on1.4.2009

Gross loanportfolio to totalassets ratio ason 1.4.2009

Yield on grossloan portfolioas on 1.4.2009

1 Bandhan 158.30 % (5) .13 % (5) 78.37 % (5) 22.16 % (5)2 Kotalipura 139.84 % (5) 1.49 % (5) 89.38 % (5) 36.10 % (3.6)3 Mahasemam 102.02 % (5) .12 % (5) 67.97 % (5) 26.67 % (4.87)4 Pustikar 141.58 % (5) .80 % (5) 80.83 % (5) 15.75 % (5)5 Sangamithra 119.13 % (5) 4.82 % (5) 100.37 % (5) 14.91 % (5)6 Sarvodaya

NanoFinance

104.72 % (5) 9.71 % (5) 87.51 % (5) 13.58 % (5)

7 SKDRDP 112.70 % (5) .31 % (5) 82.61 % (5) 13.55 % (5)8 SKS 138.88 % (5) .22 % (5) 107.01 % (5) 25.67 % (5)9 Spandana 180.04 % (5) .13 % (5) 121.64 % (5) 25.71 % (5)

Note: The values in parentheses are the scales assigned to the MFIs

the reasonable limit of 26 %. This means that though they are efficient MFIs, theydo not pass on the benefits of efficiency to clients in the form of reasonable interestrates (Table 5.5).

When the sustainability diamonds for the nine efficient MFIs were plotted allexcept Kotalipura and Mahasemam depicted an equal-sided sustainability diamond.This is represented in Fig. 5.5. As Mahasemam has a slight increase of .67 % abovethe reasonable interest rate, its sustainability diamond has only a minor deviationfrom the diamond shape. But Kotalipura’s deviation was substantial, as it had closeto 10 % increase above the reasonable interest rate of 26 %.

Among the seven efficient and sustainable MFIs, MFIs like Pustikar, Sang-hamithra, Sarvodaya Nano Finance and SKDRDP are found to be highly sociallysustainable as their yield is far lesser than the ceiling of 26 % and they still manageto reap an OSS ratio greater than 100 with this low yield. Among these MFIs,SKDRDP has the lowest yield and portfolio risk, with the highest OSS ratio andgross loan portfolio to assets ratio, making it the best performer among IndianMFIs. To ensure the credibility of the performance of these identified efficientand sustainable MFIs, the accolades and performance ratings won by them werereviewed. SKDRDP was found to have won the prestigious Microfinance IndiaAward 2010. Bandhan, Sanghamithra, Sarvodaya Nano Finance, SKDRDP, SKSand Spandana were rated by Credit Rating and Information Services of India Ltd.

Page 102: Sustainability of Indian Microfinance Institutions ||

88 5 Intermediate Participant Selection Phase: Assessment of Efficiency. . .

0

1

2

3

4

5OSS Ratio

Portfolio at Risk> 30 days Ratio

Gross LoanPortfolio toTotal Assets

Ratio

Yield to GrossLoan Portfolio

Ratio

Bandhan

Kotalipura

Mahasemam

Pustikar

Sanghamithra

SarvodayaNano Finance

SKDRDP

SKS

Spandana

Fig. 5.5 Sustainability diamond for the nine efficient microfinance institutions

Table 5.6 Microfinance institutions with their respective efficient and sustainable peers (Source:Marakkath and Ramanan 2012)

Efficient and sustainable peer MFIs Follower MFIs

Bandhan GVPustikar AWS, Janalakshmi Financial Services Pvt. Ltd.,

KBSLAB, PWMACS, RASS, SEWA BankSanghamithra Adhikar, Asomi, BFL, BISWA, BSS, Casphor MC, Cresa,

ESAF, GFSPL, GSGSK, GU, INDUR MACS,Mahashakti, NEED, RASS, RGVN, VFS

Sarvodaya Nano Finance AWS, GV, Janalakshmi Financial Services Pvt. Ltd.,PWMACS, RASS, Saadhana, SEWA Bank

SKDRDP Ujjivan, BFL, Casphor MC, JFSL, VFSSKS BASIX, BFL, BISWA, BSS, Casphor MC, ESAF, GV,

JFSL, RGVN, Sonata, Ujjivan, VFSSpandana Adhikar, Asomi, BASIX, BISWA, Casphor MC, Cresa,

ESAF, GFSPL, GSGSK, GU, GV, HIH, INDURMACS, Janalakshmi Financial Services Pvt. Ltd.,KBSLAB, Mahashakti, NEED, PWMACS, RGVN,Saadhana, SCNL, Sonata, SU, Swadhaar

(CRISIL) as high performing MFIs in the year 2009. Bandhan and SKS also featuredas the top MFIs on a global MFI rating done by Forbes for the year 2007. Pustikar,Bandhan and Sanghamithra were rated to be efficient and sustainable MFIs byQayyum and Ahmad (2006).

Thus the seven identified efficient and sustainable MFIs can be regarded as thepeers or reference group for the rest of the less efficient ones. In Table 5.6, the

Page 103: Sustainability of Indian Microfinance Institutions ||

5.7 Affect of Institutional Specific Factors on Efficient and Sustainable. . . 89

MFIs in the sample are presented against their respective efficient and sustainablepeers, as identified by CCR and BCC models under input orientation method. Thistabulation will enable the less efficient MFIs in the sample to identify those MFIs,whose managerial strategies it should understand further to enhance its efficiencyand sustainability.

The DEA model also projects the inputs that each of the lesser efficient MFIsshould have used in place of the current input values. The per cent change betweenprojected inputs and current inputs used by these MFIs shows the extent of inputminimization that these lesser efficient MFIs should achieve in order to attainefficiency. This is depicted in Table 5.7.

The less efficient sample MFIs are recommended to trim-off the inefficienciesfrom their operating structure, by taking note of the percentage reductions depictedin Table 5.7. This in turn can equip them to charge reasonable cost-covering interestrates, without passing on the cost inefficiencies to the poor clients.

5.7 Affect of Institutional Specific Factors on Efficientand Sustainable Microfinance Institutions

In order to assess if any institutional specific factors have an influence on theefficient and sustainable status of MFIs, the peculiar characteristics of the MFIswere noted. It is tabulated below in Tables 5.8 and 5.9.

Firstly a multiple regression was conducted to assess if any of the peculiarcharacteristics mentioned in the above tables—age, regulatory status, area of oper-ations, credit delivery model and size of MFI—had an influence on the efficiencyscores of MFIs. The results showed that none of these variables have an effect onefficiency of MFIs. The regression analysis was conducted using two models. Thefirst regression model treated age as a metric variable and the second model treatedit as a categorical variable (category 1: MFIs with age less than 9 years and category2: MFIs with age 9 and above). These two categories of age were used in modeltwo, as the efficient and sustainable MFIs had their starting age as 9 years.

The results of the two regression analysis undertaken for the two models aredepicted below in Tables 5.10 and 5.11, respectively.

As the regression results showed these institution-specific variables to have noinfluence on efficiency scores, a multiple discriminant analysis was undertaken toassess if these variables could discriminate between efficient and sustainable MFIsfrom the rest of the sample MFIs. The results are depicted below in Table 5.12.

The results showed that the status of these MFIs was not influenced by thesevariables. Thus it was found that though credit delivery model was an institutionalfactor which was significant in the quantitative phase, it does not have a discrimina-tory power to distinguish between efficient and sustainable MFIs from the rest of thesample. In order to confirm this finding, later in the qualitative phase of the study,the MFI managers were asked if the efficiency and sustainability status of their

Page 104: Sustainability of Indian Microfinance Institutions ||

90 5 Intermediate Participant Selection Phase: Assessment of Efficiency. . .

Tabl

e5.

7Pe

rcen

tage

chan

gebe

twee

npr

ojec

ted

and

actu

alin

puts

ofth

em

icro

finan

cein

stitu

tions

(Sou

rce:

Mar

akka

than

dR

aman

an20

12)

Con

stan

tret

urns

tosc

ale

Var

iabl

ere

turn

sto

scal

e

Seri

alnu

mbe

rM

FITo

tala

sset

sO

pera

ting

cost

per

borr

ower

Num

ber

ofcr

edit

offic

ers

Tota

lass

ets

Ope

ratin

gco

stpe

rbo

rrow

erN

umbe

rof

cred

itof

ficer

s

1A

dhik

ar�2

1.65

%�2

1.65

%�2

1.65

%�1

9.26

%�1

9.26

%�1

9.26

%2

Asm

ita0

00

00

03

Aso

mi

�20.

61%

�20.

61%

�20.

61%

�16.

31%

�64.

82%

�16.

31%

4A

WS

�28.

91%

�28.

91%

�28.

91%

�4.3

2%

�4.3

2%

�4.3

2%

5B

andh

an0

00

00

06

BA

SIX

�38.

53%

�38.

53%

�38.

53%

�37.

85%

�66.

40%

�44.

60%

7B

FL�1

6.22

%�1

6.22

%�1

6.22

%�1

6.00

%�1

6.00

%�3

0.63

%8

BIS

WA

�31.

17%

�31.

17%

�31.

17%

�30.

88%

�30.

88%

�64.

50%

9B

SS�1

8.25

%�1

8.25

%�1

8.25

%�1

6.38

%�5

7.41

%�1

6.38

%10

Cas

phor

MC

�8.3

9%

�8.3

9%

�8.3

9%

�2.6

3%

� 48.

62%

�2.6

3%

11C

resa

�25.

36%

�25.

36%

�25.

36%

�20.

44%

�24.

02%

�20.

44%

12E

SAF

�24.

01%

�24.

01%

�24.

01%

�22.

62%

�60.

18%

�22.

62%

13G

FSPL

�9.6

7%

�9.6

7%

�9.6

7%

�9.5

6%

�50.

84%

�9.5

6%

14G

SGSK

�17.

64%

�17.

64%

�17.

64%

�12.

53%

�30.

99%

�12.

53%

15G

U�1

9.49

%�1

9.49

%�1

9.49

%�1

7.45

%�1

7.45

%�1

7.45

%16

GV

�9.9

8%

�9.9

8%

�9.9

8%

�4.9

5%

�74.

19%

�4.9

5%

17H

IH�5

8.55

%�5

8.55

%�5

8.55

%�5

8.48

%�8

8.51

%�7

4.38

%18

IND

UR

MA

CS

�31.

59%

�31.

59%

�31.

59%

�23.

72%

�23.

72%

�23.

72%

19Ja

nala

kshm

iFin

anci

alSe

rvic

esPv

t.L

td.

�38.

89%

�38.

89%

�38.

89%

�36.

62%

�60.

86%

�36.

62%

20Ja

noda

ya�4

.63

%�4

.63

%�4

.63

%0

00

21JF

SL�9

.27

%�9

.27

%�9

.27

%�8

.40

%�1

1.72

%�3

2.26

%22

Kot

alip

ura

00

00

00

23K

BSL

AB

�35.

03%

�35.

03%

�35.

03%

�32.

70%

�42.

66%

�32.

70%

24M

ahas

hakt

i�1

4.05

%�1

4.05

%�1

4.05

%�3

.14

%�3

.14

%�3

.14

%

Page 105: Sustainability of Indian Microfinance Institutions ||

5.7 Affect of Institutional Specific Factors on Efficient and Sustainable. . . 91

25M

ahas

emam

00

00

00

26M

MFL

00

00

00

27N

BJK

�19.

88%

�19.

88%

�19.

88%

00

028

NE

ED

�19.

55%

�19.

55%

�19.

55%

�14.

26%

�14.

26%

�14.

26%

29N

IDA

N�7

0.87

%�7

0.87

%�7

0.87

%0

00

30Pu

stik

ar0

00

00

031

PWM

AC

S�3

3.05

%�3

3.05

%�3

3.05

%�2

0.82

%�2

0.82

%�2

0.82

%32

RA

SS�8

.72

%�8

.72

%�8

.72

%�7

.52

%�7

.52

%�7

.52

%33

RG

VN

�15.

62%

�15.

62%

�15.

62%

�15.

45%

�24.

40%

�15.

45%

34Sa

adha

na�1

1.04

%�1

1.04

%�1

1.04

%�5

.52

%�5

.52

%�5

.52

%35

Sang

ham

ithra

00

00

00

36Sa

rvod

aya

Nan

oFi

nanc

e0

00

00

0

37SC

NL

�35.

31%

�35.

31%

�35.

31%

�34.

90%

�34.

90%

�39.

61%

38SE

WA

Ban

k�4

0.63

%�4

0.63

%�4

0.63

%�3

9.94

%�6

3.26

%�3

9.94

%39

Shar

eM

AC

TS

00

00

00

40SK

DR

DP

00

00

00

41SK

S0

00

00

042

Sona

ta�3

3.80

%�3

3.80

%�3

3.80

%�3

3.48

%�7

1.02

%�3

8.99

%43

Span

dana

00

00

00

44St

ar�8

.15

%�8

.15

%�8

.15

%0

00

45SU

�4.2

0%

�4.2

0%

�4.2

0%

�26.

40%

18.1

8%

�70.

54%

46Sw

adha

ar�3

9.38

%�3

9.38

%�3

9.38

%11

.21

%�8

7.49

%5.

30%

47SW

AW

S�0

.16

%�0

.16

%�0

.16

%0

048

Ujji

van

�14.

94%

�14.

94%

�14.

94%

�7.7

2%

�66.

67%

�20.

49%

49V

FS�2

2.60

%�2

2.60

%�2

2.60

%�1

5.64

%�6

8.31

%�1

5.64

%50

WSE

�2.0

1%

�2.0

1%

�2.0

1%

00

0

Page 106: Sustainability of Indian Microfinance Institutions ||

92 5 Intermediate Participant Selection Phase: Assessment of Efficiency. . .

Table 5.8 Age, regulatory status and size of efficient and sustainable microfinance institutions

Serialnumber MFI

Age(as on 1.4.2009) Regulatory status

Size of MFI(as on 1.4.2009)

1 Bandhan 9 Regulated (NBFC) 332,462,204 (large)2 Pustikar 14 Unregulated

(cooperativesociety)

16,726,882 (small)

3 Sanghamithra 10 Unregulated(Section 25Company)

15,398,958 (small)

4 Sarvodaya NanoFinance

9 Regulated (NBFC) 19,036,739 (small)

5 SKDRDP 15 Unregulated (trust) 136,728,666 (large)6 SKS 12 Regulated (NBFC) 960,793,988 (large)7 Spandana 12 Regulated (NBFC) 787,304,262 (large)

Note: MFIs with gross loan portfolio greater than or equal to 22 million USD are regarded aslarge MFIs (Malegam Committee Report 2011)

Table 5.9 Credit model and operational area of efficient and sustainable microfinance institutions

Serialnumber MFI Credit model Area of operation

1 Bandhan Grameen Districts of Assam, Bihar, Chattisgarh, Delhi,Gujarat, Haryana, Jharkand, MadhyaPradesh, Maharashtra, Meghalaya, Orissa,Rajasthan, Sikkim, Tripura, Uttarkand,Uttarpradesh, West Bengal, Dadar and NagarHaveli (urban and rural)

2 Pustikar SHG Districts of Rajasthan, Andhra Pradesh andMaharashtra (urban)

3 Sanghamithra SHG Districts of Karnataka, Tamil Nadu and AndhraPradesh (rural)

4 Sarvodaya NanoFinance

SHG Districts of Tamil Nadu (rural)

5 SKDRDP SHG Districts of Karnataka (rural and urban)6 SKS Grameen Districts of Andhra Pradesh, Karnataka,

Maharashtra, Orissa, Madhya Pradesh, Bihar,Uttar Pradesh, Rajasthan, Uttaranchal,Himachal Pradesh, Haryana, West Bengal,Jharkhand, Chhattisgarh, Gujarat, Kerala,Tamil Nadu, Punjab and Delhi (rural andurban)

7 Spandana Grameen Districts of Karnataka, Tamil Nadu, AndhraPradesh, Goa, Orissa, Jharkhand, Chattisgarh,Maharashtra, Madhya Pradesh and Rajasthan(rural and urban)

Page 107: Sustainability of Indian Microfinance Institutions ||

5.7 Affect of Institutional Specific Factors on Efficient and Sustainable. . . 93

Table 5.10 Regression coefficients model 1

Independent variables Coefficient (t values)

Age .013 (.092)Regulatory status (NBFC, trust, society, Section 25 Company) �.147 (�.843)Area of operation (rural, urban, rural and urban) �.296 (�2.051)Credit delivery model (SHG or Grameen) �.171 (�1.105)Size of MFI (small or large) .223 (1.296)Constant 88.712 (13.576)*

Adjusted R2 .082F value 1.880N (i.e. sample size) 50

Note: Figures in parentheses show t values*Indicates values are significant at 5 % level

Table 5.11 Regression coefficients model 2

Independent variables Coefficient (t values)

Age (<9 and >9) .149 (1.065)Regulatory status (NBFC, trust, society, Section 25 Company) �.131 (�.759)Area of operation (rural, urban, rural and urban) �.284 (�1.999)Credit delivery model (SHG or Grameen) �.137 (�.878)Size of MFI (small or large) .208 (1.221)Constant 83.89 (11.869)*

Adjusted R2 .105F value 2.153N (i.e. sample size) 50

Note: Figures in parentheses show t values*Indicates values are significant at 5 % level

Table 5.12 Discriminant analysis coefficients

Independent variables

Discriminantcoefficientsmodel 1—age asmetric variables

Discriminantcoefficients model2—age as categoricalvariable (>9 and <9)

Age �.285 (.649) .681 (2.601)Regulatory status (NBFC, trust, society, Section

25 Company)�.635 (.002) �.434 (.002)

Area of operation (rural, urban, rural and urban) .658 (1.189) .639 (1.189)Credit delivery model (SHG or Grameen) .315 (.004) .491 (.004)Size of MFI (small or large) 1.021 (1.566) .806 (1.566)

Note: Figures in parentheses show F valuesNone of the values have an *. This indicates that they are not significant at 5 % level

Page 108: Sustainability of Indian Microfinance Institutions ||

94 5 Intermediate Participant Selection Phase: Assessment of Efficiency. . .

MFI was dependent on any peculiar characteristic like—age, credit delivery model,regulatory status, area of operations, size of MFI—or any other factors that theyconsidered relevant. None of the managers attributed their efficient and sustainablestatus to any factors. Instead, they attributed their efficient and sustainable statussolely to the strategies used by them in managing their MFI’s OSS. They observethat even if an MFI uses the Self-Help Group (SHG) credit delivery model whichhas higher operating costs compared to the Grameen model, through the use ofNGO-MFI partnership strategy this cost disadvantage can be mitigated. So in themanager’s view, though choice of credit delivery model is an influencing factor forsustainability, it can be strategized in such a way that it does not adversely affectefficient and sustainable status of the MFI.

5.8 Summary

The efficiency analysis and sustainability assessment undertaken in this chaptercompletes the intermediate participant selection phase of this study. The inter-mediate participant selection phase was designed in the three-phased sequentialexplanatory mixed-methods research framework to accomplish the third objectiveof this research work. Fulfilling this third objective, the study identifies the efficientand sustainable Indian MFIs who can act as reference group for other MFIs in Indianmicrofinance industry. Seven efficient and sustainable Indian MFIs are identified inthis phase using a DEA efficiency analysis model, Sustainability Diamond Modeland a benchmarking process. These seven MFIs are chosen to be the participantsfor the ensuing qualitative phase. Thus based on the findings of this phase, theefficient and sustainable MFIs, whose practices the lesser efficient sample MFIscan refer to is identified. The extent of input minimization to be achieved by eachof the lesser efficient MFIs, in order to attain efficiency, was also depicted in thisphase. Moreover, the regression analysis undertaken in this study depicts that theefficiency status of the MFIs are not dependent on institution-specific variables. Thesubsequent discriminant analysis undertaken in this phase depicts that the efficientand sustainable status of Indian MFIs cannot be discriminated on the basis of anyinstitution-specific variables. This was later confirmed by the MFI managers duringthe interview process. They attributed their efficient and sustainable status solelyto the strategies used by them in managing their MFI’s OSS. The ensuing chapterpresents a discussion on these strategies.

References

Athanassopoulos AD (1997) Service quality and operating efficiency synergies for managementcontrol in the provision of financial services: evidence from Greek bank branches. Eur J OperRes 9:300–313

Avkiran NK (2001) Investigating technical and scale efficiencies of Australian universities throughdata envelopment analysis. Socioecon Plann Sci 35(1):57–80

Page 109: Sustainability of Indian Microfinance Institutions ||

References 95

Banker RD, Charnes A, Cooper WW (1984) Some models for estimating technical and scaleinefficiencies in data envelopment analysis. Manage Sci 30(9):1078–1092

Bassem BS (2008) Efficiency of microfinance institutions in the Mediterranean: an application ofDEA. Transit Stud Rev 15(2):343–354

Berger AN, Humphrey DB (1997) Efficiency of financial institutions: international survey anddirections for future research. Eur J Oper Res 98:175–212

Berger AN, Mester LJ (1997) Inside the black box: what explains differences in the efficiencies offinancial institutions? J Bank Financ 21:895–947

Bhatt N, Tang SY (2001) Delivering microfinance in developing countries: controversies and policyperspective. Policy Stud Organ 29(2):319–334

Charnes A, Cooper WW, Rhodes E (1978) Measuring the efficiency of decision making units. EurJ Oper Res 2:429–444

Christen RP (2001) Commercialization and mission drift. The transformation of microfinancein Latin America. Consultative Group to Assist the Poor (CGAP) Occasional Paper 5,Washington, DC

Cingi S, Tarim A (2000) Performance measurement in Turkish banking system: data envelopmentanalysis Malmquist TFP index application. Turkish banks association research proceedingsseries, report no. 2000, 2

Coelli T (1996) A guide to DEAP version 2.1: a data envelopment analysis (computer) program.Centre for efficiency and productivity analysis working paper, Department of Econometrics,University of New England. http://www.une.edu.au/econometrics/cepa.htm. Retrieved 25 Dec2011

Darrat AF, Topuz C, Yousef T (2002) Assessing cost and technical efficiency of banks in Kuwait.Paper presented at ERFs 8th annual conference, Cairo

Desrochers M, Lamberte M (2003) Efficiency and expense preference behavior in Philippinescooperative rural banks. Centre Interuniversitairesur Les Risque, Les Politiques EconomiquesEt L’emploi (CIRPÉE.) Cahier de Recherché working paper series, pp 3–21

Farrell MJ (1957) The measurement of productive efficiency. J Roy Stat Soc 120:253–281Farrington T (2000) Efficiency in microfinance institutes. Microbank Bull 4:20–23Guitierrez-Nieto B, Serrano-Cinca C, Molinero CM (2007) Microfinance institutions and effi-

ciency. OMEGA Int J Manage Sci 35(2):131–142Guitierrez-Nieto B, Serrano-Cinca C, Molinero CM (2008) A DEA approach to microfinance

institutions efficiency. Paper presented at centre for European research in microfinanceseminars, Brussels

Haq M, Skully MT, Pathan S (2009) Efficiency of microfinance institutions: a data envelopmentanalysis. Asia Pac Financ Mark 17:63–97

Hassan MK, Tufte DR (2001) The X-efficiency of a group based lending institution: the case ofGrameen bank. World Dev 29:1071–1082

Micro Rate and Inter-American Development Bank (2003) Performance indicators for micro-finance institutions. A technical guide. http://microrate.com/pdf/Technical%20Guide%203rd%20Edition%20English.pdf. Retrieved 25 Dec 2011

Kalim R, Salahuddhin T (2011) Micro financing of NGOS and government: collaborative impacton poverty eradication. Inform Manage Bus Rev 2(2):81–91

Lafourcade A, Isern J, Mwangi P, Brown M (2005) Overview of the outreach and financialperformance of microfinance institutions on Africa. http://www.mixmarket.org. Accessed 25June 2010

Malegam Committee Report (2011) Report of the Reserve Bank of India sub-committee of itsCentral Board of Directors to study issues and concerns in the Micro Finance Institutions(MFI) sector. Reserve Bank of India. http://www.rbi.org.in/scripts/BS_PressReleaseDisplay.aspx?prid=23780. Retrieved 25 Feb 2010

Marakkath N, Ramanan RT (2012) Assessing the efficient and sustainable performance of indianmicrofinance institutions. Cost Manage. Thomson Reuters/RIA, 26(5):1–14

Navajas S, Schreiner M, Richard M, Claudio G, RodriguezMeza J (2000) Microcredit and thepoorest of the poor: theory and evidence from Bolivia. World Dev 28(2):333–346

Page 110: Sustainability of Indian Microfinance Institutions ||

96 5 Intermediate Participant Selection Phase: Assessment of Efficiency. . .

Nghiem H, Coelli T, Rao DSP (2006) The efficiency of microfinance in Vietnam: evidence fromNGO schemes in the north and the central regions. Int J Environ Cult Econ Soc Sustain2(5):71–78

Otero M (2000) Bringing development back to micro finance. J Micro Financ 1(1):8–19Qayyum A, Ahmad M (2006) Efficiency and sustainability of micro finance institutions in South

Asia. MPRA Paper 11674, University Library of Munich. http://www.saneinetwork.net/pdf/SANEI_VI/SANEIVI%20PROJECT%207%20Efficiency%20and%20Sustainability%20of%20Micro%20Finance%20Institutions%20in%20South%20Asia.pdf. Retrieved 25 Dec 2010

Ramanathan R (2003) An introduction to data envelopment analysis. Sage, New DelhiRobinson SM (2001) The micro finances revolution: sustainable finance for the poor. World Bank,

Washington, DCSchreiner M (2002) Aspects of outreach: a framework for discussion of the social benefits of

microfinance. J Int Dev 14(5):591–603Siems TF, Barr RS (1998) Benchmarking the productive efficiency of US banks. Financial Industry

Studies, Federal Reserve Bank of Dallas, 11–24 Dec 1998Soteriou A, Zenios SA (1999) Operations, quality and profitability in the provision of banking

services. Manage Sci 45(9):1221–1238Sufian F (2006) The efficiency of non-bank financial institutions: empirical evidence from

Malaysia. Intern Res J Financ Econ 6:1450–2887Vassiloglou M, Giokas D (1990) A study of the relative efficiency of bank branches: an application

of data envelopment analysis. J Oper Res Soc 41:591–597Woller G, Dunford C, Warner W (1999) Where to microfinance. Int J Econ Dev 1:29–64

Page 111: Sustainability of Indian Microfinance Institutions ||

Chapter 6Qualitative Phase: Management of FactorsAffecting and Discriminating Sustainability

6.1 Preface

This chapter focuses on the qualitative phase of the study, which aims to understandthe management of the factors affecting and discriminating the OSS status of IndianMFIs. The qualitative phase presents a discussion on the interviews conductedwith the managers of efficient and sustainable MFIs. Out of the seven efficientand sustainable MFIs identified in the intermediate participant selection phase ofthe study, four MFIs which were willing to participate in this study are chosenfor interviewing. The details of these four interviews undertaken in this study arediscussed in this chapter. The intent of undertaking these four interviews is tounderstand how efficient and sustainable MFIs are managing the factors affectingand discriminating the OSS status of Indian MFIs. Confirmation is also sought fromthe MFI managers about the relationships shared by these factors with OSS ratio, asobserved in the quantitative phase of the study. The results of this qualitative phaseare expected to provide guidelines to other Indian MFIs in managing their OSS.The qualitative phase will subsequently be closed by a discussion on the issues ofmismanagement of OSS and this will be dealt with separately in the next chapter.

6.2 Overview of the Qualitative Phase

As discussed in Chap. 3, the method of semi-structured interviews is used inthe qualitative phase of the study. The method of interview is chosen as it willenable the participants to freely express their views and discuss the strategies usedfor managing the OSS of their MFIs. An overview of the qualitative interviewphase is presented in this section by discussing the different steps involved in the

N. Marakkath, Sustainability of Indian Microfinance Institutions:A Mixed Methods Approach, India Studies in Business and Economics,DOI 10.1007/978-81-322-1629-2__6, © Springer India 2014

97

Page 112: Sustainability of Indian Microfinance Institutions ||

98 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

Conducting the FinalInterviews with MFIManagers at their HeadOffice

Pilot Testing the InterviewGuide: PreliminaryInterviews with ThreeMFIs & Refinement ofInterview Guide

Formulating InterviewGuide: To Explain theRelationships & theManagement of theFactors Identified in theQuantitative Phase

PreliminaryExploration of theData: By Readingthrough Transcripts &Writing Summaries

Qualitative Data Analysis& Documentation of theStrategies: Treating theFive Factors Affecting &Discriminating their OSSStatus as Codes

Triangulation: UsingQuantitative Data toCheck the Validity of theStrategies

Fig. 6.1 Interview protocol

interview protocol. Yin (1994) regards the use of a detailed protocol, as a disciplinedmethodology to increase the reliability of the qualitative research methods. Theinterview protocol used in this study is portrayed in Fig. 6.1. The steps in theinterview protocol are discussed below in the following subsections.

6.2.1 Formulating the Interview Guide

Firstly, as per the protocol, an interview guide1 is formulated in this study. Theguide comprises of questions pertaining to the five significant factors identified inthe quantitative phase of the study. The broad areas of interest in the interview guideare as follows:

(a) To understand whether the MFI managers confirm the relationships obtainedin the quantitative phase of the study, with respect to the significant factorsaffecting and discriminating the OSS status of Indian MFIs

(b) To understand how the MFIs manage the factors affecting and discriminatingits OSS status

The interview guide formulated to elicit answers from MFI managers on thesetwo broad aspects, comprised of fifteen open-ended questions.

1In the case of interviews, the questionnaire used for eliciting information from participants iscalled interview guide.

Page 113: Sustainability of Indian Microfinance Institutions ||

6.2 Overview of the Qualitative Phase 99

6.2.2 Pilot Testing the Interview Guide

After formulating the guide, it was pilot tested on three MFIs, other than MFIsincluded as participants in the study. Preliminary interviews were conducted withthe managers of these three MFIs to ensure that there is proper coherence inthe meaning of the questions included in the guide. Data obtained through thesepreliminary interviews were analysed, to ensure that the questions were capable ofeliciting the information on the management of the factors. Based on the experienceof conducting these preliminary interviews, the guide was fine-tuned and the numberof questions was reduced to twelve. This revised guide is appended at the end of thethesis (refer Appendix 3).

6.2.3 Conducting the Final Interviews

Before the conduct of the interview, in order to comply with the confidentialitynorms of the MFIs, it was agreed that the identity of the MFIs would not be disclosedin any published documents using the data collected from the MFIs. Therefore,the interviewed MFIs are proxied in this study with the names—A, B, C and D.Face to face interviews were conducted with CEOs (chief executive officers) and/orMFI managers, using the pilot-tested interview guide. It roughly took 3–4 h for aninterview. The interviews were conducted at the head office of the respective MFIs.Data collected by these interactions was recorded (in writing) by the researcher.The field notes were transcribed on the same day of the conduct of the interview toensure minimum data loss.

6.2.4 Preliminary Exploration of Data

Preliminary data exploration was done by reading through the transcripts associatedwith each of the four interviews. Transcripts were then summarized by noting downthe key points and narratives made by the MFI managers with respect to each of thefive factors addressed by the twelve questions. A summary of the qualitative datacollected during the interview is tabulated in the appendix (refer Appendix 4).

6.2.5 Data Analysis and Documentation of Strategies

Based on the transcripts and summaries, the qualitative data was analysed. Thequalitative data analysis procedure used in this study was chosen based on thethematic analysis processes discussed by Fereday and Muir-Cochrance (2006),

Page 114: Sustainability of Indian Microfinance Institutions ||

100 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

Schutz (1967), Boyatzis (1998) and Crabtree and Miller (1999). As a first stepin qualitative data analysis, coding of data was done by treating each of the fivesignificant factors as separate codes. The code definition was the same as theoperational definition of the five factors, as used in the quantitative phase of thestudy. This approach of deductive a priori selection of codes, prior to the conduct ofqualitative data analysis, is advocated by Crabtree and Miller (1999), as a means fororganization of data for subsequent analysis and interpretation. Such an a priori codeselection approach was found appropriate for this mixed-methods study, so that thefive significant factors identified in the quantitative phase can be regarded as codesfor the ensuing qualitative analysis. As the intention of conducting the interviewsare to understand how the interviewees are explaining the relations and managerialstrategies associated with these five factors, this kind of an a priori code selectionprocedure was suitable for this study. After deciding upon these factors as codes, anencoding process as outlined by Boyatzis (1998) was undertaken. Boyatzis regardsthe process of encoding as the process of organizing data for the development ofthemes from codes. In the words of Boyatzis, themes denote the pattern in datathat at minimum describes and organizes the possible observations and at maximuminterprets aspects of the phenomenon (i.e. codes). For identifying such themes, theinterview transcripts were summarized by identifying the key points discussed bythe managers. This involved several rounds of reading and memo writings doneby the researcher, based on the interview transcripts. The key points narrated bythe MFI managers with respect to each of these five codes were then treated asthemes. Thus, this data-driven thematic analysis process used in this study is inline with the inductive thematic analysis procedure discussed by Boyatzis. As theaim of this qualitative analysis procedure was not content analysis, each of the keypoints discussed by the MFI managers was regarded as a theme, by giving themsame importance as a repeated theme. As each key points was important, there wasno need felt for the use of qualitative data analysis softwares to handle the codingprocess and pattern identifications.

Each of the key points discussed by the manager was initially identified as thefirst-order themes, which represented diverse means for managing a factor. Thesefirst-order themes were then further clustered and labelled as second-order themes.Second-order themes captured the aspects discussed in the first-order theme in amore abridged form. Together, the first- and second-order themes associated witha code, denote the strategies used by different MFIs to manage a factor. Thus, thestrategies used by each of the four MFIs to manage the five factors affecting anddiscriminating their MFI’s OSS status were documented for the reference of otherMFIs operating in Indian microfinance industry. Along with the strategies, the MFImanagers also discussed some of the policy changes they considered relevant forfacilitating the management of the factors. The schematic overview of the thematiccoding process undertaken for documenting the strategies and policy suggestions,with respect to each of the five factors, is represented in Sects. 6.4.1, 6.4.2, 6.4.3,6.4.4 and 6.4.5.

Page 115: Sustainability of Indian Microfinance Institutions ||

6.3 Discussion on the Relationship Shared by the Five Factors with Operational. . . 101

6.2.6 Triangulation Using Quantitative Data

To ensure the validity of the strategies and policy suggestions, a triangulationprocess was undertaken. This occasionally required going back to the participantswho were interviewed earlier. Available supporting quantitative data collected fromthe interviewed MFIs and archival data obtained from the MIX market database wasanalysed to check the validity of the strategies. Thus, an iterative process of cyclingamong observed results of the quantitative phase, interview data of the qualitativephase and the quantitative and archival data on the 50 sample MFIs was done torefine the findings, relate them to existing microfinance literature and clarify thecontributions of the work.

Having discussed the interview protocol, the next two sections (Sects. 6.3 and6.4) discusses the two areas broadly addressed in the interview guide: relationshipand management of the factors affecting and discriminating the OSS status of IndianMFIs.

6.3 Discussion on the Relationship Shared by the FiveFactors with Operational Self-Sustainabilityof Microfinance Institutions

As discussed in Sect. 6.2.1, the MFI managers were first asked confirmatoryquestions on the relationship shared by the five factors with the sustainability ofan MFI. They were also asked to explain the relationships by mapping it on to theOSS ratio. Question numbers 2, 4, 6, 8 and 10 in the interview guide pertains to thespecific questions asked to the MFI managers to elicit this information.

The responses to the confirmatory questions are summarized in Table 6.1.As depicted in Table 6.1, all the MFI managers confirmed the relationships asso-

ciated with four factors—risk, growth, development and cost factor—as observedin the quantitative phase of the study. With respect to institutional factor, theMFIs which used the SHG credit delivery model agreed their model to be costlier

Table 6.1 Confirmation on factor relationships

Factor

MFIPortfoliorisk factor

Growthfactor

Developmentfactor

Institutionalfactor

Cost-efficiencyfactor

A * * * *

B * * * *

C * * * * *

D * * * * *

Note: *Denotes confirmation of the factor relationships as observed in thequantitative phase

Page 116: Sustainability of Indian Microfinance Institutions ||

102 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

OSS Ratio =

Operating Income

Operating Costs + Financing Costs + Loan Loss Provisions

Positive Association betweenPortfolio at Risk > 30 Daysand Loan Loss Provisions

Fig. 6.2 Mapping relationship between portfolio risk and operational self-sustainability ratio

than Grameen model. But they did not confirm the negative relationship sharedby SHG model and OSS ratio. According to them the higher group formationcosts on SHG model can be overcome by using specific deliverance mechanisms.A discussion on this mechanism is presented in Sect. 6.4, while explaining themanagerial aspects related to this factor. In this section, the explanations derivedfrom the responses of the four MFI managers on the observed factor relationshipsare discussed. The consolidated explanatory results are presented below, under thefollowing subsections.

6.3.1 Portfolio Risk Factor: Mapping the NegativeRelationship Between Portfolio Risk Greater Than30 Days and Operational Self-Sustainability Ratio

All MFI managers mapped the negative association between portfolio riskiness andOSS of MFIs by mapping an underlying positive association between an MFI’sportfolio risk greater than 30 days ratio and loan loss provision expenses. Theunderlying relationship is depicted in Fig. 6.2.

The validity of this explanation was checked by calculating the correlationbetween portfolio at risk greater than 30 days and loan loss provisions. Sample dataof 50 MFIs showed a correlation of .672 between Portfolio at Risk Greater than30 days and Loan Loss Provisions, significant at 95 %. This significant correlationvalidated the explanation.

6.3.2 Growth Factor: Mapping the Positive RelationshipBetween Gross Loan Portfolio and OperationalSelf-Sustainability Ratio

Two of the MFI managers interviewed, mapped the positive association betweengrowth and OSS of MFIs to the enhanced operating income arising on account of

Page 117: Sustainability of Indian Microfinance Institutions ||

6.3 Discussion on the Relationship Shared by the Five Factors with Operational. . . 103

OSS Ratio =

Operating Income

Operating Costs + Financing Costs + Loan Loss Provisions

Growth Shares NegativeRelationship with Costsof an MFI

Growth Shares PositiveRelationship withOperating Income

Fig. 6.3 Mapping relationship between growth and operational self-sustainability ratio

growth. The other two managers attributed it to cost reduction and scale economiesarising on account of an enlarged portfolio size. The underlying relationships aredepicted in Fig. 6.3.

The validity of these explanations was checked using the 50 sample MFIs.Sample data showed a correlation of .926 between gross loan portfolio and operatingincome, significant at 95 %. But the correlation between gross loan portfolioand operating cost, though negative, was not significant. Similar nonsignificantcorrelation existed between gross loan portfolio and total cost ratio. This depictsthat significant scale economies do not exist for Indian MFIs. This was reconfirmedby calculating the correlation between total number of borrowers and cost ratios,which again turned out to be negative but nonsignificant.

Thus, the qualitative phase proves that growth enhances sustainability of IndianMFIs mainly due to the enhanced revenue associated with it. It also shows thatthough the interviewed efficient and sustainable MFIs reported to have experiencedscale economies on account of growth, the sample MFIs were yet to reap suchsignificant scale economies.

6.3.3 Development Factor: Mapping the NegativeRelationship Between Average Loan Size PerBorrower and Operational Self-Sustainability Ratio

All the MFI managers mapped the negative association between average loan sizeand OSS of MFIs to the increased operating costs on large-sized loans. They haveexperienced the screening and monitoring costs needed per borrower to be higher,when the loan size increases. This is so because when the loan size increases, usuallythe credit officer will have to do individual loan assessments for the borrower.This makes the operating cost per borrower to be higher for larger loan sizes. Theunderlying relationship is sketched as in Fig. 6.4.

Page 118: Sustainability of Indian Microfinance Institutions ||

104 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

OSS Ratio =

Operating Income

Operating Costs + Financing Costs + Loan Loss Provisions

Average Loan Size Per Borrower Shares Positive Relationship withOperating Costs

Fig. 6.4 Mapping relationship between average loan size per borrower and operational self-sustainability ratio

The validity of this explanation was checked using the sample data. Correlationbetween log average loan size and log operating cost to gross loan portfolio ratiowas �.292, appearing significant at 95 %. But dominating this negative effect, thecorrelation between log average loan size and log operating cost per borrower of theMFI was .310, significant at 95 %.

This proves that as the managers explain, there exists a positive relationshipbetween loan size and operating cost, when it is viewed from the perspective ofper borrower.

Further we also checked whether increase in loan size had an effect on portfoliorisk. But the correlation between average loan size and portfolio risk of MFI wasnot significant.

Thus, overall there is evidence for the cost disadvantage arising on accountof increased loan size. The graphs associated with these linear relationships aredepicted in Figs. 6.5 and 6.6, respectively.

To identify the specific loan size interval at which the cost disadvantage occurs,the loan size was structured and presented along with the respective operating costsas in Figs. 6.7 and 6.8, respectively.

Though the graph shows differences in loan size around the interval of 88–176USD, an ANOVA test done to test its significance did not confirm this difference.Thus, it was not possible to identify a particular level of loan size at which thedifference was significant.

Thus, it was not possible to identify a particular level of loan size at which thedifference was significant.

But the explanation provided by the managers clarifies why the quantitativeresults showed an absence of mission drift in Indian context, contradicting thehypothesized positive association. The dominant positive association betweenaverage loan size per borrower and operating cost per borrower explains the inverserelationship shared by loan size and OSS ratio.

Page 119: Sustainability of Indian Microfinance Institutions ||

6.3 Discussion on the Relationship Shared by the Five Factors with Operational. . . 105

0.00

-0.50

-1.00

-1.50

-2.001.80 2.00 2.20 2.40 2.60 2.80 3.00 3.20

ObservedLinear

Log Average Loan Size Per Borrower

Log

Ope

rating

Cos

t to

Gro

ss L

oan

Por

tfol

io R

atio

Fig. 6.5 Negative relationship between operating cost to gross loan portfolio ratio and averageloan size per borrower

2.00

1.50

1.00

0.50

0.001.80 2.00 2.20 2.40 2.60 2.80 3.00 3.20

ObservedLinear

Log Average Loan Size Per Borrower

Log

Ope

rating

Cos

t Per

Bor

row

er

Fig. 6.6 Dominant positive relationship between operating cost per borrower and average loansize per borrower

Page 120: Sustainability of Indian Microfinance Institutions ||

106 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

Fig. 6.7 Average loan size and operating cost to gross loan portfolio ratio

Fig. 6.8 Average loan size and operating cost per borrower

6.3.4 Institutional Factor: Mapping the Negative RelationshipBetween Usage of Self-Help Group Credit DeliveryModel and Operational Self-Sustainability Ratio

All the four MFIs observed the SHG model to have a higher operating cost perborrower than Grameen model, because the group formation costs are higher in theformer model. The group formation costs and time associated with each of the creditdelivery model is depicted in Table 6.2.

Page 121: Sustainability of Indian Microfinance Institutions ||

6.3 Discussion on the Relationship Shared by the Five Factors with Operational. . . 107

Table 6.2 Group formation cost and time associated with credit models

MFIInstitutional factor(credit delivery model)

Group formationcost (in INR)

Time for groupformation

A SHG model 7,000 6 monthsB SHG model 2,500 2 monthsC Grameen model 300 35 daysD Grameen model 200 14 days

Note: Cost and time data are as reported by MFI managers

OSS Ratio =

Operating Income

Operating Costs + Financing Costs + Loan Loss Provisions

SHG Credit Delivery Model hasHigher Operating Cost Per Borrower

Fig. 6.9 Mapping relationship between self-help group credit delivery model and operational self-sustainability ratio

But as discussed in the confirmatory stage, the MFIs using SHG model had arationale for using this model and they observed that the higher group formationcosts can be overcome by using a specific deliverance mechanism. Details on thisare presented while discussing the managing strategies associated with this factor. Inthis section, the explanation for negative relationship between usage of SHG creditdelivery model and OSS of MFIs is focused on.

The underlying relationship between credit delivery model and operating costper borrower is portrayed in Fig. 6.9.

An independent sample t-test was conducted to see if there is a significantdifference in the operating cost per borrower for sample MFIs using SHG modeland Grameen Model. The results revealed MFIs with SHG model to have asignificantly higher operating cost than MFIs with Grameen model. MFIs with SHGmodel (Mean D 25.56, Std. Deviation D 22.77) reported significantly higher levelsof operating cost per borrower than the MFIs with Grameen (Mean D 14.76, Std.Deviation D 7.94); t value (2.239); degrees of freedom (48); p-value (.030). Theresults were significant at 95 % confidence interval. Further, independent sample t-tests were conducted to see if there is a significant difference in the portfolio risk andyields of MFIs, using SHG model and Grameen Model. No significant difference inrisk and yield was observed. Thus, the results reveal that the negative relationshipbetween SHG model and OSS is due to the higher operating cost associated withthis model.

Page 122: Sustainability of Indian Microfinance Institutions ||

108 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

OSS Ratio =

Operating Income

Operating Costs + Financing Costs + Loan Loss Provisions

Largest Chunk of Cost of an MFI and aDiscriminating Component of the OSS Ratio

Fig. 6.10 Mapping discriminatory relationship between cost-efficiency factor and operationalself-sustainability ratio

6.3.5 Cost-Efficiency Factor: Mapping the DiscriminatoryRelationship Shared by Operating Cost Per Borrowerto Operational Self-Sustainability Status

All the four MFI managers observe operating costs to be the highest cost componentof an MFI, with the power to discriminate the MFI’s sustainability status. Managersobserve operating cost to account for nearly 2/3 of the total cost of an MFI. Thishighest cost component, being the denominator of the OSS ratio, distinguishes thesustainability status of the MFIs. One of the managers observes:

Since the financing cost averages to around 12-13 per cent of gross loan portfolio and loanloss costs in normal conditions is around 1-2 per cent for majority of the Indian MFIs, thecrucial distinction occurs with respect to the operating costs. This is where the difference insustainability occurs.

The underlying relationship is depicted in Fig. 6.10.Having discussed the explanations provided by managers on the relationships

shared by the five factors, the next section focuses on the managerial aspect of thesefactors.

6.4 Discussion on the Management of the Five Factors

In this section, the focus is to understand how the efficient and sustainable MFIs aremanaging the five factors affecting and discriminating its OSS status. This pertainsto the second broad aspect addressed in the interview. To elicit this information,several questions were posed to the MFI managers (refer: Question numbers 3, 5, 7,9, 11 and 12 of the interview guide).

As discussed in Sect. 6.2.5, the strategies used by interviewed MFIs to managethese factors were documented using a coding process. The coding process treatseach factor as a code and each strategy as a theme associated with the code. In theensuing subsections the result of the coding process is schematically representedand each of the strategies discussed by the managers is elucidated.

Page 123: Sustainability of Indian Microfinance Institutions ||

6.4 Discussion on the Management of the Five Factors 109

Surrogate assessment& design of customer-centric products

Mitigate risk onaccount ofuncontrollable factors

Portfolio at Risk > 30days ratio & Write-OffRatio < 10 per cent

Equip poor to indulgein income generatingactivities

Talk to Defaulter,Enforce Joint Liability,Write-Off

Take action againstCoercive RecoveryPractices

Training CreditOfficers

Providing CapacityBuilding Services

Providing InsuranceCoverage

Tracking PortfolioIndicators

Following UpDefaulting Loans

AppointOmbudsman

ManagingPortfolio Risk

Factor

First Order Theme Second Order Theme Deductive Code

Fig. 6.11 Thematic coding process for the deduced code: portfolio risk factor

6.4.1 Portfolio Risk Factor: Strategies and Policy Suggestions

The strategies and policy suggestions discussed by the MFI managers with respectto portfolio risk factor are enumerated and explained in this section. A schematicrepresentation of the coding process undertaken to arrive at these strategies andpolicy suggestions is depicted in Fig. 6.11. In Fig. 6.11; portfolio risk factor denotesthe code and each of the emerging themes denote the associated strategies and policysuggestions.

The MFI managers observed prevention of portfolio risk to be critical forensuring loan recovery performance. This is so, because once the loans becomedelinquent, MFIs, which operate on uncollateralized lending models, will have lesscontrol over its recovery. Nevertheless, they also discussed about risk detective andcorrective strategies. The managers also made suggestions to regulator for facilitat-ing the management of this factor. Based on these observations, the discussions onrisk factor are categorized as (I) Preventive Strategies, (II) Detective and CorrectiveStrategies and (III) Suggestion to Regulator.

I. Preventive Strategies

(i) Training Credit Officers: The MFI manager who advocates this prac-tice explained the importance of adopting risk-preventive strategies that

Page 124: Sustainability of Indian Microfinance Institutions ||

110 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

emphasize on training of MFI credit officers. According to him trainingcredit officers to do surrogate assessment of client’s creditworthiness isessential as it can facilitate in the design of customer-centric products. Inhis view a product is called customer-centric if it matches with the client’sfinancial needs and repayment capacity. Provision of such customer-centric products which are in tune with the client’s cash flow patterns is ameans to prevent default on loans. In the words of the manager:

Training credit officers to do income assessment based on surrogates is a goodstrategy to prevent default on loans. For instance, if a client has a buffalo as anincome generation source, then the credit officer should know how to assess theworth of this asset as a surrogate of the client’s income. The loans should bedesigned by matching the repayments with the expected cash flows from thisasset. This has ensured low portfolio risk for us and the recoveries of loans hasalways happened smoothly. No exploitative practices have ever been used by ustill date for recovery.

To check the validity of this strategy, a causal relationship betweentraining expenses and risk could not be tested using regression analysis.This was because the 50 sample MFIs either did not incur specifictraining expenses incurred per credit officer in surrogate assessment andcustomer-centric product design or was reluctant to share such data.Nevertheless, we managed to collect data on the training expenses incurredper credit officer,2 from the four efficient and sustainable MFIs and fromfour other MFIs in the sample which had the highest portfolio risk. Anindependent sample t-test was conducted to examine whether there isa significant difference between the training expenses per credit officerincurred by efficient and sustainable MFIs and other sample MFIs withhigh portfolio risk. The test revealed a statistically significant differencebetween the two groups. Efficient and sustainable MFIs (Mean D 1401.75,Std. Deviation D 181.29) reported significantly higher levels of trainingexpenses than the less efficient and unsustainable MFIs (Mean D 625.5,Std. Deviation D 126.01); t value (7.032); degrees of freedom (6); p-value(.000), significant at 95 % confidence interval. Thus, it was found that thefour less efficient and unsustainable MFIs with high portfolio risk incurredless training expenses per credit officer than the efficient and sustainableMFIs. As the latter group observes such training to positively contributetowards portfolio quality, it is recommended that the less efficient andunsustainable MFIs take note of this risk preventive strategy.

(ii) Providing Capacity Building to Clients: Two of the MFI managersinterviewed claimed that they could reduce the vulnerability to default,by providing capacity building services to clients. Business development

2The usage of training expenses as a surrogate variable to validate this strategy has somelimitations. This is so because if credit officer is already well trained in surrogate assessment andclient-centric product design, he/she will not need further training. This will skew data for MFIswith younger credit officers that need more training.

Page 125: Sustainability of Indian Microfinance Institutions ||

6.4 Discussion on the Management of the Five Factors 111

services, social welfare services and financial literacy services were thecapacity building services provided by these MFIs. One of the managerswho used this strategy reported:

Capacity building can equip the clients to indulge in sustainable microenter-prises, without defaulting on the loans taken from MFIs.

The definitions given by the manager for each of the components ofcapacity building services are as follows:

(a) Business Development Services: Services that can help the poor toimprove their business performance, e.g. training in business skills forentrepreneurs

(b) Social Welfare Services: Non-financial services that can improve thequality of life of the clients, e.g. health and hygiene camps andeducational services

(c) Financial Literacy Services: Services that can educate the poor aboutpersonal financial management, which in turn can help them to be abetter customer of the MFI

To test the validity of this strategy, the causal relationship betweenportfolio at risk and provision of such services was tested on the 50 sampleMFIs, using a simple regression analysis. The results were not significant.An independent sample t-test was conducted to see if there is significantdifference in portfolio risk of the sample of 50 MFIs, with respect toprovision of welfare services. The results showed no statistical evidenceto prove that provision of welfare services reduced risk. This couldprobably be attributed to the quality differences in the services providedby the lesser efficient sample MFIs when compared to the efficient andsustainable MFIs. For instance, a survey through the archival informationcollected from the website of the lesser efficient sample, MFIs showed thatmost of the MFIs, which we included as providers of welfare services inthe analysis, have not reported to have imparted financial literacy services,in the manner offered by interviewed MFIs. The interviewed efficient andsustainable MFIs rendered financial literacy services, such that it educatedthe poor about their repayment capacity and the financial discipline theyare expected to adhere to, in order to ensure repayment. This according tothem contributes positively to their portfolio quality. This aspect was miss-ing in the welfare services provided by most of the lesser efficient MFIs.

Thus, overall though this strategy was not validated by the entiresample data owing to our inability to capture the quality differences in theservice provision, the managerial experience was that those clients whomade use of business development services and financial literacy serviceswere less prone to default.

(iii) Providing Insurance Coverage to Clients: One of the MFI managersobserved that defaults arising on account of uncontrollable and unexpectedfactors like death, health issues, accidents and crop failure can be reducedby providing insurance coverage to the clients.

Page 126: Sustainability of Indian Microfinance Institutions ||

112 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

Table 6.3 Portfolio riskindicators

MFI Write-off ratioPortfolio atrisk > 30 days

A 4.28 4:82

B 0.29 5:00

C 0.00 :13

D 0.67 :13

To test the validity of this strategy, the causal relationship betweenportfolio at risk and provision of insurance services was tested on the50 sample MFIs, using a simple regression analysis. The results werenot significant. An independent sample t-test was conducted to see ifthere is significant difference in portfolio risk of the 50 MFIs withrespect to provision of insurance. Though the MFIs which providedinsurance had a lower mean risk than the MFIs which did not provideinsurance, the difference was not statistically significant. The test resultscannot be completely relied on to check the validity of this strategy,because the MFIs which were included in the category of insuranceproviders were not homogenous with respect to the insurance schemesoffered. Majority of the less efficient sample MFIs offered only lifeinsurance schemes, whereas the interviewed MFI which emphasized onthe provision of insurance services offered a wider range of services likemedical insurance, livestock insurance and crop failure insurance.

II. Detective and Corrective Strategies

(iv) Tracking Key Indicators of Portfolio Risk: In addition to the use ofpreventive strategies against risk, one of the managers emphasized on theneed for risk monitoring. According to the manager, constant and simulta-neous monitoring of two risk indicators is crucial for an MFI—portfolioat risk greater than 30 days and write-off ratio. These two indicatorsneed concurrent monitoring, because an MFI’s impressive portfolio atrisk greater than 30 days ratio should not be due to a high write-offsassociated with its loans. Risk can be said to be in control only if boththese indicators are low. Otherwise, it is a sign of portfolio riskiness whichneeds managerial attention. According to the manager:

Ideally these two risk indicators should be 10 per cent or below, to ensure controlof portfolio risk within acceptable parameters.

Table 6.3 shows that these two risk indicators were below 10 % for thefour interviewed MFIs.

(v) Following-Up Defaulting Loans and Taking Corrective Action: One of theMFI managers interviewed discussed the steps to be taken when a clientdefaults.

When a weekly payment is missed, the credit officer must follow-up thisclient immediately. The officer must inquire the reason for non-payment

Page 127: Sustainability of Indian Microfinance Institutions ||

6.4 Discussion on the Management of the Five Factors 113

and discuss with the defaulter the possibility for repayment and theconsequences of delinquency. As delinquency will prevent further loandisbursements to the group, the client must be asked to make good thedefault. If the client still does not arrange for making the payment, thenthe joint-liability principle should be enforced on the group members. Thegroup members should be asked to jointly pool in the defaulter’s share. Ifthe group members refuse to assume the joint liability, then the loan willhave to be written off. Thus, the manager points out:

Apart from prompt follow-ups and enforcement of joint-liability, there is veryless control over delinquent loans. Beyond this, an MFI should not try to ensurerecovery by the use of coercive loan recovery practices.

III. Suggestion to Regulator

(vi) Appointment of Ombudsman to Handle Client Complaints: Since over-bearing and coercive collection practices at the credit officer-client inter-face have alleged concerns about client protection in Indian microfinanceindustry, one of the managers suggested that RBI should set up a well-functioning client complaint redressal mechanism to address this issue.This should be akin to that of an ombudsman that operates in Indianbanking industry. As per the views of this manager, presence of such asystem at central level is essential to ensure that MFIs are managing theirrecoveries without exploiting the clients.

Though the manager suggested for the presence of an Ombudsman byRBI, we felt that operating such a system at central level would havepractical difficulties as there will be too many MFI borrowers who canpotentially lodge a complaint. When this apprehension was discussed withthe manager, he said that this can be addressed by using local resourcessuch as Panchayat or District Magistrate as the first level ombudsman.

Thus, having discussed the different strategies and policy suggestions formanaging portfolio risk factor, the next section focuses on how the efficient andsustainable MFIs were managing the growth factor.

6.4.2 Growth Factor: Strategies and Policy Suggestions

The strategies and policy suggestions discussed by the MFI managers with respectto growth factor are enumerated and explained in this section. The schematicrepresentation of the coding done for growth factor is depicted in Fig. 6.12. InFig. 6.12, growth factor denotes the code and each of the emerging themes denotesthe associated strategies and policy suggestions.

MFI managers observed that frantic growth strategies, without considering itsimpact on the cost and risk of MFIs, would not contribute to the sustenance of theMFI in the long run.

Page 128: Sustainability of Indian Microfinance Institutions ||

114 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

Cost involved in expansion tonew areas supported bycapital: Horizontal Growth

Market Penetration inexisting markets: increasecredit officer productivity,standardized products

Replication of SuccessfulBranch Staff Training,Customized Products &Monitoring Growth

Role of Credit OfficerProductivity in balancinggrowth and risk

Prevent adverse effect ofgrowth: Multiple Borrowing

Deciding GrowthStrategy

Achieving VerticalGrowth

Achieving HorizontalGrowth

Balancing Growthwith Portfolio Risk

Formation of CreditInformation Bureau

ManagingGrowthFactor

First Order Theme Second Order Theme Deductive Code

Fig. 6.12 Thematic coding process for the deduced code: growth factor

One of the managers explained how his MFI plans for growth, by choosingits growth strategy—(vertical and horizontal growth strategies) in relation to thecapital base available to support the costs involved in growth. Since the MFIs haveexperienced a positive relationship between credit officer productivity and portfoliorisk, they also advocated the use of growth-balancing strategies that maintainoutreach at acceptable levels of portfolio risk. The managers also made suggestionsto the regulator to address the adverse effects of an MFI’s growth operations.Based on these observations, the discussion on growth factor is categorized asfollows: (I) Growth-Enhancing Strategies, (II) Growth-Balancing Strategies and(III) Suggestion to Regulator.

I. Growth-Enhancing Strategies

(i) Understanding the Costs Involved in Expansion and Deciding the GrowthStrategy: One of the MFI managers, whose MFI has operations spreadover both North and South India, discussed how he went about decidinghis organization’s growth strategy. His discussion emphasized on howcrucial it is for an MFI manager to plan for growth. He described howinitially the MFI started its operations in South India and then graduallybegan to think of expansion plans to the North. While planning forthis transition and growth, the managers calculated the cost involved inhorizontal expansion. The estimates of costs involved in setting up ofnew branches, new product development and hiring and training of new

Page 129: Sustainability of Indian Microfinance Institutions ||

6.4 Discussion on the Management of the Five Factors 115

staff were assessed. Similarly, the cost of installation of the infrastructureneeded to support and monitor growth was assessed. The capital baseavailable to support this additional cost was then ascertained. The MFIdelayed their expansion plans, until the managers were confident that therewas a large capital base to support the increased costs. Therefore, themanager says:

Horizontal growth needs capital support to contribute to the sustainability of theMFI. If the managers feel such a support isn’t present at the moment, then theyshould stick on to vertical expansion plans in familiar markets.

To check the validity of this statement, there was no available datato assess how an MFI decided the compatibility between estimatedcosts and capital base of the MFIs. But with the available data onthe capital of the MFIs and their growth strategy (vertical or hor-izontal), we checked whether all the seven efficient and sustainableMFIs identified in this study actually had a larger capital base to sup-port its horizontal growth than that at which an MFI with verticalgrowth operates. An independent sample t-test was conducted to seeif there is a significant difference in the capital base of the sevenefficient and sustainable MFIs with respect to their growth strategy.The results revealed MFIs with horizontal growth strategy to have asignificantly higher capital base than MFIs with vertical growth. Efficientand sustainable MFIs with horizontal growth (Mean D 656447416, Std.Deviation D 236959175) reported significantly higher levels of capitalthan the MFIs with vertical growth (Mean D 55822678, Std. Devia-tion D 73173114); t value (�4.908); degrees of freedom (5); p-value(.004), significant at 95 % confidence interval. Similarly another indepen-dent sample t-test was conducted to see if there is a significant differencein the capital base of the seven less efficient and unsustainable MFIsidentified in the study, with respect to their growth strategy. The resultsrevealed no significant difference in the capital base of the MFIs withrespect to their growth strategy. Thus, no statistical evidence was foundamong the MFIs for the choice of growth strategy being made based oncapital support.

(ii) Achieving Vertical Growth through Market Penetration in Existing Mar-kets: An MFI with intense vertical operations in South India remarkedthat this has been possible for them due to their exclusive focus onfamiliar demographies. Based on the assessment of the client needsthrough market research, the MFI has been providing standardized loanproducts which cater to the needs of the vast majority of clients in thatdemography. Though such standardized product delivery was the usualpractice, variations in the basic products in accordance with customerrequirements was also done as and when required. But according to theMFI manager, the emphasis on standardization has helped them to enhancethe credit officer productivity of their MFIs (the credit officer productivity

Page 130: Sustainability of Indian Microfinance Institutions ||

116 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

Table 6.4 Portfolio risk andcredit officer productivity MFI

Portfolio atrisk > 30 days

Credit officerproductivity

A 4:82 1,306B 5:0 1,731C :13 522D :13 503

for these MFIs are 1,306 and 1,731, respectively). This has also resultedin market penetration and intensive growth of operations for the MFIs.

(iii) Achieving Horizontal Growth through Replication of Successful Branch:The MFI manager who have operations spread across the length andbreadth of the nation opines that horizontal growth can be achieved byreplicating successful branch models in new markets. The new marketselection is described by the manager as follows:

We select area where there is potential for branch replication. Target areasusually have 240 households such that there is minimum of 60 client potentialin a village and minimum of 3000-4000 client potential in a slum.

The manager observes that while expanding operations in new markets,the existing MFI staff should train the newly hired staff for a briefinterval, until they are comfortable to handle the new branch operationsindependently. Under the guidance of existing staff, they should betrained to design customized products for the new market. The cost andrisks associated with the new operations needs to be monitored and themanagers advocate the use of a Management Information System (MIS)to serve this purpose.

II. Growth-Balancing Strategies

(iv) Balancing Growth with Portfolio Risk—Understanding the Role of CreditOfficer Productivity: In the view of an interviewed MFI manager, irre-spective of the growth strategy adopted by an MFI, the managers shouldensure a balance between the growth of operations and portfolio risk of theMFI. The manager observes the presence of a direct relationship betweencredit officer productivity and portfolio risk. When the credit officersincrease their caseload of borrowers, the quality of credit assessment andrelationship building with clients can suffer. So an MFI aiming to achievegrowth by enhancing credit officer productivity must be wary of thisdownside risk. He remarks:

Any effort to pursue growth (quantity) by compromising on portfolio quality willnot contribute to sustainability of the MFI.

To check the validity of the above observations, we first glanced at thequantitative data on the associated parameters, which is tabulated below asin Table 6.4.

Page 131: Sustainability of Indian Microfinance Institutions ||

6.4 Discussion on the Management of the Five Factors 117

Table 6.5 Portfolio riskregressed on log borrower percredit officer

Independent variable Coefficient (t values)

Log borrower per credit officer .389 (2.925)*

Constant �11.939 (�2.467)*

Adjusted R2 .134F-value 8.558*

N (sample size) 50

Note: Figures in parentheses show t values*Indicates values are significant at 5 % level

The data showed that MFI with higher credit officer productivity hadcomparatively higher risk. To check the existence of this association on alarger sample of 50 MFIs, the correlation between portfolio at risk greaterthan 30 days ratio and borrower per credit officer was ascertained. Theresults depicted a positive correlation of .322, significant at 95 % level.When portfolio at risk was regressed on log borrower per credit officer, theresults showed that a 100 % increase in borrower per credit officer wouldincrease risk by .389. Table 6.5 depicts the results of this regression.

Thus, the results validate the existence of a downside risk, whichis small in magnitude but significant. As MFI’s sustainability is highlydependent on its portfolio recoveries this result is relevant for MFImanagers. To further investigate on this and to examine at what level ofcredit officer productivity a significant difference in portfolio risk exists, anindependent sample t-test was conducted on the 50 MFIs. The test resultsshowed significant difference at the level of 550 borrowers per credit offi-cer; t value (�2. 366); degrees of freedom (48); p-value (.022), significantat 95 % confidence interval. MFIs with credit officer productivity less than550 (Mean D 1.185, Std. Deviation D 1.435) reported significantly lowerlevels of portfolio at risk >30 days than the MFIs with credit officerproductivity of 550 or above (Mean D 4.2873, Std. Deviation D 7.545).This results show that MFI managers enhancing credit officer productivityhave to be wary of this downside risk operating, when caseload of creditofficers are increased above 550 borrowers.

III. Suggestion to Regulator

(v) Formation of a Credit Information Bureau: One of the MFI managersobserves overleveraging of clients to be an adverse effect of MFI’s growthstrategies. But MFIs have little control over this problem as clients seldomreveal their correct leverage information and there is no credit historyavailable to verify what they report. MFIs usually become aware of theoverleveraging problem only when the client faces defaults and becometrapped in multiple borrowings. Therefore, the manager suggests theregulator to form and maintain a centralized credit information bureauto which all MFIs shall be made to compulsorily report their portfoliolending details. Such a database will enable MFIs to check the extent

Page 132: Sustainability of Indian Microfinance Institutions ||

118 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

Creates tendency to borrowfrom moneylenders &other MFIs

Small loan size does notfulfill client’s growingneeds as they graduate

Provide loans in a mannerthat suits client’srepayment capacity andfinancial needs

Design flexible loanproducts matching client’sfinancial needs

Small Loan Size AloneDoes Not Result in

Development

Small Loan Size Resultsin Multiple Borrowing

Progressive Increase inLoan Size Results inTrue Development

Removal of Cap onAverage Loan Size

ManagingDevelopment

Factor

First Order Theme Second Order Theme Deductive Code

Fig. 6.13 Thematic coding process for the deduced code: development factor

of multiple borrowings taken by the clients from different MFIs. TheMalegam Committee Report on Indian Microfinance (2011) promisesthe formation of such a credit information bureau. The MFI managerinterviewed called for the establishment of such a system at the earliest.

Thus having discussed the different strategies and policy suggestions for man-aging growth factor, the next section focuses on how the efficient and sustainableMFIs were managing the development factor.

6.4.3 Development Factor: Strategies and Policy Suggestions

The strategies and policy suggestions discussed by the MFI managers with respectto development factor are enumerated and explained in this section. The schematicrepresentation of the coding done for development factor is depicted in Fig. 6.13.In Fig. 6.13, development factor (average loan size per borrower) denotes the codeand each of the emerging themes denotes the managerial experiences and strategiesassociated with it.

All the managers interviewed negated the theoretical belief that development forthe poor can be attained by providing small average loan size per borrower. Theyobserve that when loan size is limited to small amounts, the financial needs of clientswould remain partially unfulfilled, making them resort to multiple borrowings frominformal sources. According to the MFIs managers, true development orientationlies in progressively providing larger-sized loans that are in tune with the financial

Page 133: Sustainability of Indian Microfinance Institutions ||

6.4 Discussion on the Management of the Five Factors 119

Table 6.6 Top five sourcesthat poor prefer to meet theirnon-routine expenditure(Source: Johnson and Meka2010)

Source

Share of non-routineexpenditure financed througha given source (in per cent)

Loan from friends/relative 43

Own income or savings 29

Loan from moneylender 13

Loan from landlord 11

Loan from MFI/SHG 6

Note: Since the poor have relied on more than one source,the percentage entries in the table are independent of eachother

needs and repayment capacity of the poor. The managers therefore suggest thatthe regulators should remove cap on loan size imposed on Indian MFIs. Based onthese observations the discussion on development factor is categorized as follows: I.Managerial Experience, II. Recommended Strategy and III. Suggestion to Regulator.

I. Managerial Experience

(i) Small Loan Size Alone Does Not Result in True Development: Majority ofthe poor client’s financial needs during the initial loan cycles is small. Butgradually their needs increase. One of the managers speaks out:

Our experience is that when the client gradually recovers out of poverty, they willbe in need of larger sized loans, which is well within their repayment capacity. Ifthe MFI does not provide it, they once again fall prey to moneylenders and otherinformal source, defeating the purpose of development. MFIs will not be able toretain such clients.

Thus, MFIs negate the theoretical belief that providing loans of smallsize alone results in serving the poor.

When we asked why the poor could not resort to formal financialinstitutions like banks for large-sized loans, the MFI managers said thatnon-price barriers like elaborate documentation and income assessmentsmake them reluctant to do so.

One of the managers adds to this by stating that this is in particularfor the non-productive and non-routine financial needs of the clients.Clients may need non-productive loans on a short notice to meet expensesrelated to marriages, festivals and medical grounds. The MFI based in thedistrict of Andhra Pradesh cites the results of a research work done onthe microfinance clients in their area to support this claim. The results ofthe research done by Johnson and Meka (2010) (Centre for Microfinance,Institute for Financial Management and Research) report the top fivesources that the poor prefer to meet their non-routine expenses. Non-routineexpenses in this research pertain to non-productive expenditure related tohealth, marriage, festival, buying of agricultural inputs, home improvement,repair and reconstruction. The results of this research are depicted below inTable 6.6.

Page 134: Sustainability of Indian Microfinance Institutions ||

120 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

The manager cites the low share of MFIs in this research and states that,if MFIs could meet the larger financial needs of the poor, it would not loseits clients to other informal players like money lenders. According to himthe limited loan size of MFIs defeats the developmental mission of reachingthe poor. He states thus:

It is unfortunate that poor approach informal sources and money lenders whocharge interest rate in the range of 24-120 per cent per annum, when MFIs wereoperating in the range of 20-40 per cent.

(ii) Small Loan Size Induces Tendency for Multiple Borrowing Among Clients:One of the MFI managers observes that the poor resort to multipleborrowing to meet their unfulfilled financial needs. He is of the opinionthat multiple borrowing from different financial intermediaries who donot understand the cash flow pattern of the clients would result in over-indebtedness making the clients unable to repay their loans. This problemof multiple borrowing is alleged to have caused client suicides, leading to amicrofinance crisis in India. According to one of the interviewed managers,the crisis proves that MFI’s true development orientation lies in designingloans in amounts that match the client’s financial needs and repaymentcapacity, rather than partially fulfilling their needs and making them go formultiple loans.

II. Recommended Strategy

(iii) Progressively Increase Average Loan Size Based on Client’s Needs: Asthe MFI managers were of the opinion that small loan size need notnecessarily connote true development for the poor, they progressivelyincreased the loan size based on the client needs. We did not check thevalidity of the strategy by surveying the clients view on this aspect. As ourstudy is more institution oriented, we checked the validity of this strategyby collecting data on the client retention rate of the MFIs. The four MFImanagers were asked to calculate their client retention rate for the year2009, based on the following formula:

Client retention rate D 1 � Drop-out rate

Where, Drop-out rate is the number of clients who completed a loanin the 6 months prior to the year 2009 and did not take out a subsequentloan in that period or within 1 month following the end of the 6-monthperiod, divided by the total number of clients who completed a loan in the6 months prior to the end of the year 2009. Table 6.7 shows the associateddata.

As the client retention rate was higher for MFIs with larger averageloan size per borrower, it can be taken as an evidence for the higher client-satisfaction level at this level.

Page 135: Sustainability of Indian Microfinance Institutions ||

6.4 Discussion on the Management of the Five Factors 121

Table 6.7 Average loan sizeper borrower and clientretention rate

MFI

Development factor(average loan size perborrower in USD)

Client retentionrate (in per cent)

A 130 68B 129 55C 144 89D 215 96

III. Suggestion to Regulator

(iv) Remove Cap on Average Loan Size: Currently, RBI imposes the followingcaps on MFI loans:

(a) Maximum loan size of INR 35,000 (first cycle) and INR 50,000(subsequent cycles)

(b) Maximum borrower total indebtedness of INR 50,000

MFI managers request the regulator to remove these caps on loan sizeas it will induce the tendency for multiple borrowings among the clients.According to the MFIs only a flexible loan size that matches the repaymentcapacity and financial needs of the clients will result in true development,not mere provision of small-sized loans.

Thus having discussed the managerial experiences, strategies and policy sugges-tions for managing development factor, the next section focuses on how the efficientand sustainable MFIs were managing the institutional factor.

6.4.4 Institutional Factor: Strategies and Policy Suggestions

The strategies and policy suggestions discussed by the MFI managers with respect tothe institutional factor, credit delivery model, are enumerated and explained in thissection. The schematic representation of the coding done for institutional factor isdepicted in Fig. 6.14. In Fig. 6.14, institutional factor (credit delivery model) denotesthe code and each of the emerging themes denotes the managerial experience andstrategies associated with it.

In the case of the institutional factor (credit delivery model), the MFI managersagreed that cost-wise Grameen model is more economical than SHG model, as thegroup formation cost is lower for the former. Though there was a consensus on this,two MFI managers, who used SHG credit delivery model, were of the opinion thatthis cost disadvantage could be minimized by adoption of partnership models. Theypreferred the usage of the costlier SHG model to Grameen model, as the formerdevotes more time for empowering the poor than the latter model.

Moreover, they observe that it is possible to reduce the higher group formationcosts of SHGs by entering into NGO-MFI partnerships. So in the view of these

Page 136: Sustainability of Indian Microfinance Institutions ||

122 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

Grameen Model has lowerGroup Formation Cost &Time Compared to SHGmodel

SHG model spends more timefor group formation andnurturing of clients

SHG group formation costreduced by outsourcing groupformation task

Grameen morecost effectivethan SHG

SHG model suitablewhen client’s

need empowerment andnot just credit

NGO-MFIPartnerships

ManagingInstitutional

Factor

First Order Theme Second Order Theme Deductive Code

Fig. 6.14 Thematic coding process for the deduced code: institutional factor

managers, with the use of NGO-MF I partnership strategy, both the benefits ofcost advantage and client empowerment can be combined in the SHG model. Basedon these managerial observations, the discussion on the institutional factor (creditdelivery model) is categorized as recommended strategies.

I. Recommended Strategies

(i) Low-Cost Grameen Model for Sustainability Enhancement: The MFI man-agers who were interviewed observed that the operating cost on Grameenmodel is lesser when compared to the SHG model. This is so because theyhave experienced the group formation cost and time to be lesser for theformer model. The quantitative figures put forth by the managers to supporttheir observations on group formation were cited earlier in this chapter inTable 6.2. The table shows one of the MFIs to have formed their clientgroup in 14 days. As this was the minimum number of days taken, weenquired about the process involved in such group formation. The processis documented in Table A5.1 of the appendix.

The results of the two independent sample t-tests, which checked if thereis significant difference between the operating cost, portfolio risks and yieldof both these models, were discussed in Sect. 6.3.4. The test results on thesample of 50 MFIs confirmed the Grameen model to have less operating costthan the SHG model, with no significant risk and yield differentials. Beingthe low-cost model, the managers using Grameen model recommended itsusage to enhance the sustainability of the MFI.

(ii) Self-Help Group Model Recommended when Empowerment of Clients AreNeeded: Though the MFI managers who use SHG model agree their modelto have more operating cost, they rate and recommend their model to bemore appropriate for dealing with clients for whom credit is not the onlymissing link to development. According to these managers, the SHG model

Page 137: Sustainability of Indian Microfinance Institutions ||

6.4 Discussion on the Management of the Five Factors 123

MFI SHG

4. Disburse Loans to SHG

5. Makes Repayments to MFI

1. Forms &NurturesSHGs

2. Links SHGsto MFIs

NGO: FormsSHG & Links with

MFI

3. M

FI

Pay

sC

omm

ission

to N

GO

s

Fig. 6.15 Non-governmental organization-microfinance institution partnership model (Source:Marakkath 2012)

spends more time on nurturing the group and empowering the clients thanthe Grameen model, thereby equipping the clients to indulge in sustainableincome-generating activities. One of the managers reported:

We empower the poor by training the client on income-generating activities, savingsand book keeping, making them socially empowered to manage their financialneeds. We recommend the use of SHG model when client empowerment and socialinclusion is needed in addition to financial inclusion. We believe our customerneeds both.

(iii) Use of Non-Governmental Organization-Microfinance Institution Partner-ship Model to Reduce Cost of Self-Help Group Formation3: The MFImanagers, who use the SHG model, observed that it is possible to reducethe higher group formation costs of SHGs by entering into NGO-MFIpartnerships. The MFI partners with NGOs and outsources the groupformation and nurturing activities to NGOs at a nominal cost. The managerstates:

We pay 350 INR per linkage or in special cases 1 percent of the loan amount lentto the NGOs as commission. The actual cost of group formation comes to 7000INR per group. Cost savings for us on account of this partnership is around 6650INR per group.

The NGO-MFI partnership model is depicted below in Fig. 6.15.As shown in Fig. 6.15, the MFI enters into a partnership with NGO,

whereby the latter forms the SHGs and links it to the former. The MFIpays a commission to the NGO for facilitating and undertaking this groupformation task. The NGO usually have affinity groups of poor affiliated

3This material is published by the author in the article: Marakkath N (2012) Innovative strategiesused by Indian MFIs to achieve cost efficiency. Int J Financ Bank Stud 1(1):2147–4486.

Page 138: Sustainability of Indian Microfinance Institutions ||

124 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

to them for activities related to its own welfare mission and thereforethey normally do not have to put in additional efforts to form SHGs.The NGO interacts with the group members on a day-to-day basis andtherefore inculcating financial discipline among the members is done hand-in-hand with their normal activities. The commission, received from theMFIs for undertaking this group formation and nurturing services, serves asan additional income for the NGO, though it is usually a minimal amount.From the NGO’s perspective, by linking the SHGs with MFI, they are ableto address the capital constraints faced by their poor clients who undertakeincome-generating activities. This advantage acts an incentive for the NGOto form SHGs, without compromising on the financial discipline of themembers. Identifying such NGOs, which have a motivation to enter to thispartnership, is crucial for the success of the model.

From the MFI’s perspective, this partnership relieves them from under-taking the group formation and nurturing activities, thereby enabling themto concentrate more on its core activity of financial intermediation. Thus inthe NGO-MFI partnership model, MFI, which otherwise incurs INR 7,000on group formation, outsources this task to NGO for a nominal fee of INR350. This results in a saving of INR 6,650 for the MFI. Therefore in thismodel, the MFI merely lends loans to the SHGs, which are already formedand nurtured by the NGO. The SHGs then directly makes the repaymentof the loans to the MFIs. The repayment of loan by the group membersis not the NGO’s responsibility. But in the MFI’s experience, since loandelinquencies adversely affect the NGO’s chances to sustain the capitalsupport for their clients, the NGOs takes special care to ensure quality andfinancial discipline of the groups formed. Therefore, the MFI has neverexperienced this partnership to adversely affect its portfolio quality.

Thus having discussed the different strategies and policy suggestions for manag-ing credit delivery model, the institutional factor, the next section focuses on howthe efficient and sustainable MFIs were managing the cost-efficiency factor.

6.4.5 Cost-Efficiency Factor: Strategies and Policy Suggestions

Though operating cost per borrower was the variable used in quantitative phase ofthe study, during the interviews, the managers were asked to discuss the strategiesused for managing both operating costs and financing costs.

Therefore, the schematic representation of the coding done for cost-efficiencyfactor had two figures, namely, Figs. 6.16 and 6.17, one each for operating costs andfinancing costs.

Based on the interactions with the MFI managers, the discussions on cost-efficiency factor are categorized as (I) Strategies for Reducing Operating Costs, (II)Strategies for Reducing Financing Costs and (III) Suggestion to Regulator.

Page 139: Sustainability of Indian Microfinance Institutions ||

6.4 Discussion on the Management of the Five Factors 125

Means for reducinggroup formation cost

Recruit local staff whoshare client’s householdeconomics

Monitor loan portfolio, maintain data integrity, transparency and co-ordinate

Maintain over-night cashat the minimum Cash Flow

Management

Increase CreditOfficer

Productivity

Use of IT-enabledManagement

Information System

NGO-MFIPartnerships

ManagingOperating

Costs

First Order Theme Second Order Theme Deductive Code

Fig. 6.16 Thematic coding process for the deduced code: operating costs

Injection of LargeAmount of Capital atLower Costand Fast Recovery of Loans

Cost efficient MFIs can negotiate with investors and donors to obtain lowcost funds

Efficient MFIs rewardedby permitting belowbase rate financial rates

Deposit a source offinance if rating normson deposits arerelaxation

Relax DepositNorms

Lower operationalcost resulting inlower financing

costs

Reward CostEfficiency

Securitization ofloans

ManagingFinancing

Costs

First Order Theme Second Order Theme Deductive Code

Fig. 6.17 Thematic coding process for the deduced code: financing costs

Page 140: Sustainability of Indian Microfinance Institutions ||

126 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

Table 6.8 Borrower percredit officer and operatingcost per borrower

MFIBorrower percredit officer

Operating cost perborrower (in USD)

A 1,306 4

B 1,731 3

C 522 16

D 503 10

Table 6.9 Operating costregressed on credit officerproductivity

Independent variable Coefficient (t values)

Log borrower percredit officer

�.337 (2.483)*

Constant 2.067 (5.567)*

Adjusted R2 .095F-value 6.167*

N (sample size) 50

Note: Figures in parentheses show t values*Indicates values are significant at 5 % level

I. Strategies for Reducing Operating Costs

(i) Non-Governmental Organization-Microfinance Institution Partnership forReducing Group Formation Cost: The strategy of outsourcing groupformation and nurturing tasks to NGOs for a nominal commission wascited by one of the MFI managers as a means to reduce their operatingcosts. The details of this strategy were discussed earlier in this chapter, inthe context of reducing the group formation cost of SHGs and thereforeare not restated.

(ii) Increasing Productivity of Credit Officers: All the four MFIs interviewedstate that by increasing the productivity of credit officers in credit deliveryprocess, cost reduction can be achieved. This was confirmed by theassociated data provided by the MFI managers, as shown below inTable 6.8.

One of the managers cited that by recruiting staff who share thesame household economics of the poor clients, credit assessment andrelationship building can be easily achieved, which in turn enhancescredit officer productivity. The negative relationship between credit officerproductivity and operating cost per borrower was also validated by the 50sample MFIs. When log operating cost per borrower was regressed on logcredit officer productivity, it was depicted that a 100 % change in creditofficer productivity will reduce operating cost per borrower by 33.7 %.The results of this regression are summarized below in Table 6.9.

But as noted earlier in this chapter, (refer Table 6.5) when portfolio atrisk was regressed on log borrower per credit officer, the results showedthat a 100 % increase in borrower per credit officer would increase risk by.389. So this cost reduction strategy also has a downside risk associated

Page 141: Sustainability of Indian Microfinance Institutions ||

6.4 Discussion on the Management of the Five Factors 127

with it. Though this is small in magnitude, since recoveries constitute asignificant aspect of MFI’s sustainability, MFI managers should be waryof this aspect.

(iii) Usage of Information Technology (IT)-Enabled Management InformationSystem (MIS): All the four MFIs use back-end MIS at their branches,which are updated by the on-field information captured by credit officer’susage of point of sale (POS) technology. This according to them augmentsthe MFI’s operational efficiency. Personal digital assistants (PDAs) areused by credit officers to record the on-field transactions. After everytransaction, a printed receipt is issued to the customer. Thereafter, the datapertaining to the transactions are transferred by the credit officer into theMIS, by connecting the PDA to the computers located at the branch. Thus,it reduces the need for manual data entry and helps to manage informationflow by generating timely reports. In the words of one of the managers,the advantages of having an MIS are as follows:

The use of MIS is crucial for monitoring the portfolio quality of the large numberof small loans that we handle. MIS helps to gain a 360 degree view of ouroperations, which in turn helps us to perpetually monitor all the five factorsyou are talking about.

The manager’s discussion on the impact of MIS on the five factors issummarized below in Fig. 6.18.

The MFI managers observe that IT-enabled MIS has helped them toreduce close to 5 % of their operating cost. The managers emphasizedthat the use of MIS and PDAs does not require huge fixed cost andinfrastructural support that is usually associated with networked POSand mobile-based channels. Delphix and MIFOS were some of the MISsoftwares used and recommended by the interviewed MFI managers. Themanagers cited MIS to facilitate information flow and transparency in MFIoperations. According to one of the managers:

Based on the daily collection data compiled and updated by the credit officer atthe branch level MIS, financial statements are generated at the head office. ThusMIS facilitates information based management at the MFI level and compliancewith the information reporting requirements at the regulatory level.

(iv) Cash Flow Management: To attain operational efficiency, cash manage-ment techniques are used by MFIs. Managing the treasury in such amanner that cash never remains idle is a technique used for effectiveutilization of cash flows. The manager who advocated this strategycommented:

By maintaining cash flows in such a way that the overnight cash held isminimum, we ensure perfect circulation of funds. Cash is collected and disbursedwithin a week for further on-lending.

Page 142: Sustainability of Indian Microfinance Institutions ||

128 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

MIS

CostEfficiency

Factor

DevelopmentFactor

InstitutionalFactor

PortfolioRisk Factor

GrowthFactor

Manages Growth:Monitors growth of largenumber of small loansalong with their repetitiverepayments

Manages Client -interface& FacilitatesDevelopment: Design loansize in accordance withclient needs & repaymentcapacity

Manages Credit Delivery:Manages data, informationflow and field operations bygenerating timely reports

Manages Risk: Monitorsportfolio quality by trackingrisk indicators on perpetualbasis

Manage Costs: Cost of operations& products are monitored, so as todebottleneck inefficient processes.Facilitates transparency in pricingloans

Fig. 6.18 Information-based management with management information system

II. Strategies for Reducing Financing CostsIn the case of financing costs, MFI managers observe that it is almostuncontrollable as cost of funds always averages around 12–13 % for majorityof MFIs in the industry. It is the rate at which MFIs source funds from banks.Nevertheless, the following strategies are used by the interviewed MFIs toreduce their cost of funds:

(v) Securitization4: Reserve Bank of India permits only NBFC MFIs to usethe securitization refinance option. One of the NBFC MFIs uses this routeto reduce their cost of finance. Manager of the MFI which uses this methodsays:

Injection of a large amount of capital at lower costs is possible throughsecuritization. We have 22.22 million USD worth loan portfolio securitized with

4This material is published by the author in the article: Marakkath N (2012) Innovative strategiesused by Indian MFIs to achieve cost efficiency. Int J Financ Bank Stud 1(1):2147–4486.

Page 143: Sustainability of Indian Microfinance Institutions ||

6.4 Discussion on the Management of the Five Factors 129

MFI Borrowers1. Issue of Loans to Clients

4. Repayment of Principal & Interest

Banks

PrioritySector

Lending

3.Pay

s C

ash

2.Transfer P

ool of Assets

5. Pass on collectionPeriodically

Investors forSecuritizedInstruments

9. Pass on Collection Periodically

8. Pays Cash

6. Priority Lending

7. Issue of Securities

Fig. 6.19 Securitization process (Source: Marakkath 2012)

bank. It allows us to liquidate the loans before they actually mature and therebyobtain cheap funds to make room for fresh asset creation. The cost of funds onsecuritized deal is approximately 8.75 per cent. This is far lower than the normalcost of funds for MFIs which is at 12 to 13 per cent.

The securitization process as explained by the manager is presentedbelow in Fig. 6.19.

As shown in Fig. 6.19, after issuing loans to the clients, the MFItransfers the loans to banks interested in a securitization deal. The bankwhich purchases the pool of assets then pays back cash at a discountedrate of interest to the MFI. The MFI will continue to service the sold outloans on behalf of the bank and will pass on the collections periodicallyto the bank. The MFI will be financially responsible for any losses onthe sold out loans, up to a certain percentage as agreed at the time of thesecuritization contract. This clause is termed as first loan default guaranteein the contract.

Just as the MFI gets its loans liquidated, the bank too has an advantagein entering in such a deal. The bank can use this purchased loans tofulfil their priority sector lending requirements. It can also pool theseassets and redistribute it as securities to new investors. For the investor,securitized microfinance loans are attractive as it mature much faster thanother industry investments. The maturity period ranges from 6 months to 3years and portfolio quality is generally high on microfinance loans. Thus,securitization is a win-win deal for all the parties involved. But since thereis no active secondary market for securitized microfinance instruments,usually the banks either use it to meet their priority lending requirements

Page 144: Sustainability of Indian Microfinance Institutions ||

130 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

or resell it to other banks that face the similar need. So if the redistributionelement in this model is not there, then it becomes a mere portfolio buy-out model between the MFI and the bank, with no issue of securities. Suchportfolio buy-out model was used by the non-NBFC MFIs. So, either thesecuritization or portfolio buy-out model can be used by MFIs to reducetheir cost of funds.

(vi) Reducing Operating Cost to Reduce Cost of Financing: MFI A says thatby being cost-efficient in its operations, it has been able to negotiate withfinancial institutions and donors to get funds at lower rate. According tothe manager of MFI A, being an MFI which has low cost of operations andwhich charges low interest rates, it has been able to attract considerableamount of low cost of funds. Its cost of funds average around 6.03 %,when the normal cost of funds for the industry amounts to 12–13 %.

III. Suggestion to Regulator

(vii) Recognize and Reward Cost Reduction: One of the MFI managersobserved that since interest rates and margins are caped for Indian MFIs(at 26 % and 12 % respectively with effect from 2011), there wouldbe no incentive for MFIs to reduce cost and to charge a lower interestrate. Therefore to incentivize MFIs to reduce cost, he suggests thatthe least cost players in Indian microfinance market, who charge thelowest interest rates, should be rewarded by the regulator. They must bepermitted to source funds from banks at below base rates (i.e. below theminimum interest rate charged by banks in disbursing loans). This wouldbe an encouragement for all MFIs to attain cost-efficiency and chargelower interest rates than the present ceiling of 26 %.

(viii) Relax Norms for Deposit Mobilization: One of the MFI managers sug-gested that the deposit mobilization norms for MFIs should be relaxed,as deposits can serve as additional source of funding for an MFI’s onlending and investments.

Indian MFIs have several restrictions on deposit mobilization. Currentregulation stipulates only NBFCs and Cooperatives (only from members)to accept deposits. NBFCs can currently accept deposits only if it obtainsan investment grade rating. Since rating agencies consider uncollateral-ized MFI portfolios to be risky, none of the Indian MFIs have obtained aninvestment grade rating. Therefore, the manager suggests that the ratingnorms should be reframed to be in tune with the characteristics of anMFI, which operates on uncollateralized lending models.

Thus having discussed the strategies and policy suggestions for managing thecost-efficiency factor, the next section draws the conclusion for this chapter.

Page 145: Sustainability of Indian Microfinance Institutions ||

6.5 Summary 131

6.5 Summary

The qualitative inquiry undertaken in this chapter throws light on how the efficientand sustainable MFIs are managing the five factors affecting and discriminatingtheir OSS. Prior to the discussion on the strategies used to manage the five factors,confirmation was sought from the MFI managers about the observed relationshipsshared by the factors with OSS ratio. All the MFI managers confirmed the relation-ships associated with four factors—risk, growth, development and cost factor—asobserved in the quantitative phase of the study. The underlying relationships whichconfirm the observed relationships in the quantitative phase were as follows:

(i) The negative relationship shared by portfolio at risk greater than 30 days(portfolio risk factor) with OSS was attributed to the positive association thatthe former shares with loan loss provisions.

(ii) Gross loan portfolio (growth factor) shared a positive association with OSS, asit had a positive association with operational income. Its negative associationwith costs was not significant.

(iii) SHG credit delivery model (institutional factor) shared a negative associationwith OSS, as its group formation costs were higher compared to Grameenmodel. But users of the former model were of the view that this higher costcould be mitigated using NGO-MFI Partnerships.

(iv) Average loan size per borrower (development factor) shared a negative rela-tionship with OSS, as it had a positive association with operating cost perborrower. This was so because the screening and monitoring costs were higheron larger-sized loans.

(v) Operating cost per borrower (cost-efficiency factor) shared a discriminatoryrelationship with OSS, as it is the major variable component of the OSS ratioof an MFI. Since the other components of the OSS ratio, namely, revenue,financing cost and loan loss costs come within a fixed range for most MFIs,the major distinguishing factor is its operating costs.

The strategies used by the MFIs to manage these five factors affecting anddiscriminating their OSS status were of interest and this was then understoodthrough the interview process. An elaborate discussion on the each of the strategiesused for managing the five factors affecting and discriminating the OSS of MFIs ispresented in this chapter. This section summarizes these strategies and presents itschematically using a fishbone diagram as shown below in Fig. 6.20. The diagramportrays attaining OSS as the effect. Managing the five factors using the abovediscussed strategies is depicted as the causes contributing to this effect. As seenin Fig. 6.20, the portfolio risk factor is found to be managed by adopting riskpreventive strategies that emphasize on training of MFI credit officers and provisionof insurance coverage and capacity building services to clients.

Page 146: Sustainability of Indian Microfinance Institutions ||

132 6 Qualitative Phase: Management of Factors Affecting and Discriminating. . .

Managing GrowthFactor

Managing CostEfficiency Factor

ManagingInstitutional

Factor

ManagingDevelopment

Factor

ManagingPortfolio Risk

Factor

AttainingOSS

*Deciding Growth StrategyBased on Capital Base* Balancing Growth withRisk

Managing Financing Costs*Securitization*Negotiating Based on LowOperating Cost

Managing Operating Costs*NGO-MFIPartnerships*Enhance Credit Officer Productivity*IT enabled MIS*Cash flow Management

*Grameen cost effective*SHG more empowering*NGO-MFI partnerships toreduce group formation costs

*ProgressivelyIncrease LoanSize

*Training Credit Officers,*Insurance Coverage,*Capacity Building Services

*Monitoring Risk Parameters*Following Up Loans Defaulters

Fig. 6.20 Attaining operational self-sustainability by managing five factors

Risk detective and corrective strategies that emphasize on following-up delin-quent loans and monitoring risk parameters, like portfolio at risk greater than30 days and write-off ratios, are also used by the MFIs. With respect to thedevelopment factor, the MFIs managers interviewed observed that they are unable tomanage the development factor, using small average loan size per borrower, as theclient’s financial needs remain unsatisfied at this level. According to the MFIs, truedevelopment orientation lies in progressively providing larger loans size that meetsthe financial needs and repayment capacity of the poor. Credit delivery model, theinstitutional factor, is found to be managed by MFIs based on the model’s cost-effectiveness and applicability to the client’s empowerment needs. Based on theparameter of cost-effectiveness, the MFIs recommend the use of Grameen model.But the SHG model which has higher operating cost is still preferred by MFIs dueto its potential to contribute to social inclusion. Its usage is recommended for clientsfor whom credit is not the only missing link to development. By entering into NGO-MFI partnership models, the MFIs are found to overcome the cost disadvantage onSHG model. Growth factor is found to be managed by growth-enhancing strategiesthat result in vertical and horizontal expansion and growth-balancing strategies thatbalances outreach with risk. The MFIs also discuss cost-reduction strategies thatreduce their operating cost and financing cost. Operating costs are reduced mainlyby increasing credit officer productivity. But since this method has a downside riskassociated with it, it is recommended that the managers be conscious of the samewhile using this strategy. The usage of an IT-enabled MIS is also cited as a means toreduce operational inefficiencies and is recommended as a means to monitor cost,growth and risk parameters concomitantly. Entering into NGO-MFI partnershipsis a means cited by MFIs using SHG model to reduce their operating cost. The

Page 147: Sustainability of Indian Microfinance Institutions ||

References 133

role of cash management is also emphasized as a means to reduce operating costs.With respect to reducing financing cost, the MFIs experience their status of beinga low-cost player to have helped them to source funds at a lower rate than themarket average. Securitization practices that can refinance loans and inject freshcapital at a lower rate of interest are also used by MFIs. Apart from these strategies,the MFI managers also put forth suggestions to the MFI regulator to facilitate themanagement of the five factors, without exploiting the poor. Formation of clientcomplaint redressal mechanism, credit information bureau, removal of caps ofaverage loan size per borrower, relaxation of norms for deposit mobilization andrewarding cost-efficient MFIs are the suggestions made by MFIs to regulators. Thevalidity of the strategies and suggestions are checked using available supportingquantitative figures associated with each of these factors. The managers of the lesserefficient MFIs are recommended to refer to these strategies while managing theirMFI’s OSS.

References

Boyatzis R (1998) Transforming qualitative information: thematic analysis and code development.Sage, Thousand Oaks

Crabtree B, Miller W (1999) A template approach to text analysis: developing and usingcodebooks. In: Crabtree B, Miller W (eds) Doing qualitative research. Sage, Newbury Park,pp 163–177

Fereday J, Muir-Cochrane E (2006) Demonstrating rigor using thematic analysis: a hybridapproach of inductive and deductive coding and theme development. Int J Qual Method 5(1):7

Johnson D, Meka S (2010) Access to finance in Andhra Pradesh. Centre for Microfinance. Institutefor Financial Management and Research Publication, Chennai

Malegam Committee Report (2011) Report of the Reserve Bank of India Sub-Committee of itsCentral Board of Directors to Study Issues and Concerns in the Micro Finance Institutions(MFI) Sector. Reserve Bank of India. http://www.rbi.org.in/scripts/BS_PressReleaseDisplay.aspx?prid=23780. Retrieved 25 Feb 2010

Marakkath N (2012) Innovative strategies used by Indian MFIS to achieve cost efficiency. Int JFinanc Bank Stud 1(1):2147–4486

Schutz A (1967) The phenomenology of the social world (trans: Walsh G, Lehnert F). NorthWestern University Press, Evanston (Original German Work Published, 1932)

Yin RK (1994) Case study research: design and methods, 3rd edn. Sage, Thousand Oaks

Page 148: Sustainability of Indian Microfinance Institutions ||

Chapter 7Qualitative Phase: Mismanagementof the Factors Affecting and DiscriminatingSustainability—Learnings from IndianMicrofinance Crisis

7.1 Preface

The qualitative phase of the study is wound up in this chapter, by pursuing a dis-cussion on the mismanagement issues of OSS. The preceding chapters of this bookpresented an analysis on sustainability management issues and it pertained to theperiod 2005–2009. It was undertaken with the premise that attaining sustainabilityis pertinent for all MFIs to perpetually operate in the sector. Though this premiseholds good, the dangers involved in being overly conscious about sustainability,without taking into account its impact on client welfare cannot be overlooked.Our discussion on sustainability would remain incomplete unless we reflect on thisaspect. The dangers of client exploitation in the name of sustainability can happenif there is a mismanagement of the determinants and discriminants of sustainability.From 2010 onwards when a crisis hit the Indian microfinance industry, there waswidespread allegations about such sustainability mismanagement. Therefore in thischapter, using the microfinance crisis as reference, the dangers involved in themismanagement of factors affecting and discriminating sustainability are discussed.

7.2 Indian Microfinance Crisis: A Learner’s Perspective

In the year 2010, a crisis hit Indian microfinance industry, at a time when one ofIndia’s largest MFIs, SKS, went for an initial public offering (IPO) (Kumar andRozas 2010). Though until then SKS was regarded to be one of the best playersof Indian MFI industry, there was widespread media allegations, stating its IPOlaunch, to have marked the onset of a time when Indian MFIs are losing focus onsocial goals. There was no conclusive evidence to the media reports of the distress-driven client’s suicides that backed these allegations. Bamzai (2010) reports, VikramAkula, the CEO of SKS to have responded to these allegations as follows: ‘In the13 cases of SKS borrowers committing suicides, each one of them had a reason

N. Marakkath, Sustainability of Indian Microfinance Institutions:A Mixed Methods Approach, India Studies in Business and Economics,DOI 10.1007/978-81-322-1629-2__7, © Springer India 2014

135

Page 149: Sustainability of Indian Microfinance Institutions ||

136 7 Qualitative Phase: Mismanagement of the Factors Affecting and Discriminating. . .

unrelated to our microfinance lending – ranging from harassment from in-laws todomestic disputes with a husband. None of the SKS borrowers who committedsuicide were defaulters. There is no basis of blaming them upon collection’.Thus SKS refrained from assuming responsibility of these media allegations andinvestigations are still on this matter.

But however the entire episode led to a crisis and resulted in a completereputation loss for the industry, attracting regulatory attention in the form ofinterest rate ceilings and loan restrictions (Malegam Committee Report 2011). Manyinstitutions were alleged to be masquerading as MFIs and profiting at the expense ofthe poor. Sustainability thus began to be largely considered as a negative term thatis grossly mismanaged by the MFIs.

Amidst this crisis scenario, in their article ‘Help Microfinance, Don’t Kill it’,Abhijit Banerjee, Pranab Bardhan, Esther Duflo, Erica Field, Dean Karlan, AsimKhwaja, Dilip Mookherjee, Rohini Pande and Raghuram Rajan exhorted the sectorby discussing that the need of the hour is to restore the poise of the industry.They observed that microfinance like any other industry is not without problems.Being an emerging industry, there are bound to be frictions between MFIs, clients,existing moneylenders, political players and government schemes, as an enablingregulatory environment to monitor microfinance operations is yet to evolve in India.But at this critical juncture when the industry is growing and yet to mature, if theattention gets stuck with these frictions, some of which are genuine and some ofwhich are motivated by the vested interests of market players, it will only amountto hamper the growth of the industry. The industry till date has one of the bestfinancial intermediation solutions to offer to the poor, and therefore reviving itto regain its potential and lost hope is imperative. So in this study, the authordoes not spend time analysing the veracity behind the allegations that resulted ina crisis but, nevertheless, reviews literature to see what MFIs need to take carein order to ensure that their sustainability pursuits are not at the cost of clientwelfare. This can be considered to be one of the key learnings that the crisis hasto impart. The crisis clearly depicts how attempts made to achieve sustainability,if done oblivious of client welfare, would prove dangerous for the very existenceof MFIs. This danger often occurs when the MFIs, knowingly or unknowingly aremismanaging the determinants and discriminants of sustainability. Mismanagementis a scenario where an MFI merely works on these factors for augmenting itsfinancial returns, unaware of how this would impact its client welfare. This scenarioof mismanagement would doom the death knell for the MFIs or for that matter of anysocial enterprise, as it would be a clear case of mission drift. This chapter discussesthese mismanagement issues.

7.3 Discussion on the Mismanagement of the Five Factors

Sustainability is pertinent for microfinance, but if the MFIs get obsessed withattaining sustainability, oblivious of its social goal, then it can undermine thevery spiritual foundation of this industry that aims for poverty alleviation through

Page 150: Sustainability of Indian Microfinance Institutions ||

7.3 Discussion on the Mismanagement of the Five Factors 137

financial inclusion. An exploration into Indian microfinance crisis, undertaken inthis chapter by citing various literature sources, corroborates this. It clearly depictsthe dangers involved in mismanagement of the factors affecting and discriminatingthe sustainability of MFIs. A discussion on this aspect is presented below.

7.3.1 Portfolio Risk Factor: Over-Indebtedness and CoerciveRecovery Practices

Portfolio risk factor is captured by the PAR>30 days ratio in this study. MFIs whenworking for sustainability and portfolio quality can have the tendency to aim at rapidexpansion of their operations with a zero tolerance for delinquency rates (Prabhu2011). Rapid expansion of operations will amount to lending indiscriminatelyresulting in over-indebtedness among the poor clients, which can make them mostvulnerable for default and distress-driven suicides. This goes against the interestof the client’s welfare. Similarly by aiming to attain zero nonperforming assetsstatus, the MFIs can have the tendency to maintain a high borrower to creditofficer ratio, in some cases going up to 700 per credit officer.1 Consequently,the relationship between the borrower and credit officer weakens, with the latterhaving no clue about the cash flow patterns of the former. The officers, whohave no background knowledge about the clients in their operational area, maytend to indulge in overbearing behaviour with the poor clients. As their incentivesare tied with achieving zero delinquency rates in their operational area, they willhave a justification for the use of such overbearing behaviour for recovery. Thisattempt to maximize the performance of the loans will amount to mismanagementof the portfolio risk factor for augmenting short run sustainability. Crisis showsthat such short-sighted strategies would raise concerns about clientele protectionin the industry. Therefore frantic growth achieved through comprising the qualityof borrower and client officer relationships, would mean long term sustainabilitychallenges for the MFIs (Nadiya et al. 2012).

7.3.2 Growth Factor: Expansion Plans and Investor Pressures

Growth factor is captured in this study by the gross loan portfolio of MFIs. Thetendency seen among MFIs to capture a burgeoning microfinance market witha growing loan portfolio has been cited in the media as the root cause of themicrofinance crisis in India. It is observed that when hectic growth and expansionplans are pursued by MFIs, it may call for the use of commercialized funds fromequity and debt investors (CGAP 2010). Such funds will often invite pressures

1In our discussions in Chap. 6 we have seen that the borrower to credit officer ratio crossing 550can result in downside portfolio risks

Page 151: Sustainability of Indian Microfinance Institutions ||

138 7 Qualitative Phase: Mismanagement of the Factors Affecting and Discriminating. . .

to meet the return expectations of debt and equity investors, as if it were anyother commercial investment. These return expectations can tend to influence thebehaviour of their credit officers adversely, if the MFI managers also get drivenby the investor pressures. Levy of exorbitant interest rates, multiple lending andcoercive loan recovery practices may then tend to be the means to meet the growthand return expectations of the owners and investors (Nadiya et al. 2012). All thesepractices though would earn short-run profits for MFIs would soon deviate themfrom its social goal. The reputation risks resulting from such mission drift wouldhamper the future growth of the MFI. These problems associated with growth canbe rectified only if MFIs have a clear vision to achieve steady growth, with the rightbacking of patient investors who are socially inclined for the cause they invest in.Thus it is imperative for the MFI managers to get the definition of scale/growthright. Scaling up should mean gradual and steady efforts put in to increase impactfor its clientele. Any relentless pursuit of growth, warranting the usage of ‘capitalwithout conscience’ and requiring the MFI to succumb to ‘investor pressures’ wouldmean death and not growth.

7.3.3 Development Factor: Multiple Borrowings and ClientSuicides

The development factor that turned out significant in this study is average loan sizeper borrower. As discussed in the qualitative phase of the study, mere provisionof small loan size will not result in development orientation. Sinha (2010) alsomakes a similar observation. By limiting loans to small amounts, an MFI mayreduce its own portfolio risk at the individual level, but it will fail to fulfil theborrower’s financial needs. This will make them approach other MFIs and moneylenders, leading to the problem of multiple borrowing. Such multiple borrowingfrom different financial intermediaries who do not understand the cash flow patternof the clients will result in over-indebtedness making clients unable to repay theirloans. This problem of multiple borrowing is alleged to have caused client suicides,leading to a microfinance crisis in India. Thus the crisis proves that MFIs truedevelopment orientation lies in designing loans in amounts that match the client’sfinancial needs and repayment capacity, rather than partially fulfilling their needsand making them go for multiple loans (Nadiya et al. 2012).

7.3.4 Institutional Factor: Supply Driven Credit Model

Institutional factor is captured in this study by the credit delivery model of the MFI.The pre-ccisis period saw rapid scaling-up with the usage of the existing credit

Page 152: Sustainability of Indian Microfinance Institutions ||

7.4 Summary 139

delivery models of microfinance. Thought these models aids outreach, Guna andMarakkath (2013) observes that microfinance can have high relevance for its clientsif it uses a bottom-up demand driven model for its product design and delivery. Butin most of the existing models, all processes, right form group formation to productdesign and development, are supply driven. The problem with such a supply drivenmodel, which is more conducive for scalability and breath of outreach, is that itwould have less scope to identify the needs of the clients. This happens because theclients are not at the centre of the model. Formulating a model, that has clienteleperspective at its core, is critical to make microfinance more beneficial and welfare-oriented for the poor.

7.3.5 Cost-Efficiency Factor: Strained Customer Relations

Cost-efficiency factor is captured in this study by the cost per borrower ratio ofthe MFI. Sinha (2010) observes that the scale efficiencies enjoyed by Indian MFIscould be a cause for worry, if it is the result of its frantic expansion operationsachieved by disproportionately increasing the borrowers to credit officer ratio. Thisis so because such frantic growth may exacerbate the portfolio risk of the MFI inthe long run. This possibility of a down side risk was cited in the analysis donein the qualitative interviews too. Therefore efficiency achieved by straining therelationship between borrowers and credit officers of the MFIs may tend to instigatethe usage of coercive practices for loan recovery as seen during the crisis time.Thus the cost-efficiency achieved through hectic expansion plans can also exploitthe poor clientele. Therefore care needs to be taken to ensure that efficiency is notat the expense of client welfare (Nadiya et al. 2012).

7.4 Summary

The term ‘mismanagement’ is used in this chapter to denote a situation whereMFIs are managing the factors affecting and discriminating its OSS, in such away that it maximizes its financial sustainability, without considering its impacton clientele welfare. This is dangerous for the Microfinance Institutions and theindustry as a whole, as it works against the tenet guiding the sector—sustainabilityis only a means to achieve the goal of poverty alleviation, and not an end in itself(Rhyne 1998). Using Indian microfinance crisis as a reference, in this chapter, thedangers in sustainability mismanagement are discussed. This discussion is expectedto serve as a guidance to prevent the dangers imminent in pursuing sustainabilityindiscriminately, oblivious of client welfare.

Page 153: Sustainability of Indian Microfinance Institutions ||

140 7 Qualitative Phase: Mismanagement of the Factors Affecting and Discriminating. . .

References

Bamzai (2010) We’ve never been accused of abetting suicides. Business TodayBanerjee A, Bardhan P, Duflo E, Field E, Karlan D, Khwaja A, Mookherjee D, Pande R, Rajan R

(2010) Help microfinance, don’t kill it. Indian ExpressConsultative Group to Assist the Poor (CGAP) 2010 Indian microfinance goes public: the SKS

initial public offering. http://www.microfinancegateway.org/gm/document-1.9.47761/FN65-1.pdf. Retrieved 25 Jan 2010

Guha, Marakkath (2013) Are microfinance services relevant for micro-entrepreneurial growth: acase based discussion from Indian microfinance models. Paper accepted for presentation at4th international conference on Institutional and Technological Environment for Microfinance(ITEM4), Paris

Kumar V, Rozas D (2010) Exclusive: SKS microfinance journey to IPO—an insidestory. microfinance focus. http://www.microfinancefocus.com/news/2010/05/17/exclusive-sks-microfinance-journey-to-ipo-an-insidestory/. Retrieved 25 Jan 2010

Malegam Committee Report (2011) Report of the Reserve Bank of India sub-committee of itscentral board of directors to study issues and concerns in the Micro Finance Institutions(MFI) sector. Reserve Bank of India. http://www.rbi.org.in/scripts/BS_PressReleaseDisplay.aspx?prid=23780. Retrieved 25 Feb 2010

Nadiya M, Olivares-Polanco F, Radha Ramanan T (2012) Dangers in mismanaging the factorsaffecting the operational self-sustainability (OSS) of Indian microfinance institutions (MFIs) –an exploration into Indian microfinance crisis. Asian Econ Financ Rev 2(3):448–462 (AsianEconomic and Social Society)

Prabhu G (2011) Publically held microfinance firms are a form of organized crime in emergingeconomies. Paper presented at academy of management meeting, cross divisional paper sessionof managing in emerging economy and multinational contexts, San Antonio

Rhyne E (1998) The Yin and Yang of microfinance: reaching the poor and sustainability.MicroBank Bull 2(1):6–8

Sinha S (2010) How to calm the charging bull: an agenda for CGAP in the decade of the Teneeis.Micro-Credit Rating International Limited, Gurgoan

Page 154: Sustainability of Indian Microfinance Institutions ||

Chapter 8Summary of Findings, Implicationsand Conclusion

8.1 Preface

This research work was undertaken with the aim of understanding the issues relatedto sustainability management among Indian MFIs. This broad research aim wasaccomplished in this study by pursuing four specific research objectives. A three-phased sequential explanatory mixed-methods research design was formulated tofulfil these objectives. The preceding three chapters were a discussion on the three-phased mixed-methods analysis undertaken in this study, to achieve these fourobjectives. This chapter begins by summarizing these three analytical phases. Inthe first phase of the study, which is quantitative in nature, the objective was toidentify the factors that affect and discriminate the OSS status of Indian MFIs. Amultiple regression model, which tests 11 hypothetical relationships, was used toidentify the factors affecting the OSS of Indian MFIs. The results of the model,when tested on a sample data of 50 Indian MFIs over the period 2005–2009,inferred four significant factors that Indian MFI managers must concentrate onto enhance the OSS of their organizations—growth factor (gross loan portfolio),portfolio risk factor (portfolio at risk greater than 30 days), institutional factor(credit delivery model) and development factor (average loan size per borrower).These four significant factors affecting the OSS of MFIs and the two componentsof OSS ratio—namely, revenue generation factor and cost-efficiency factor—werethen used as lagged independent variables in a multiple discriminant analysismodel, to assess their predictive power in discriminating the OSS status of IndianMFIs. The results of the model indicate cost-efficiency factor (operating cost perborrower) to be the single significant discriminator of the OSS status of IndianMFIs. The model was validated using validation data sets and it showed at least90 % prediction accuracy during all of these estimations. Thus, by undertaking themultiple regression and discriminant analysis, the quantitative phase identified fivesignificant factors determining and discriminating the OSS status of Indian MFIs.

N. Marakkath, Sustainability of Indian Microfinance Institutions:A Mixed Methods Approach, India Studies in Business and Economics,DOI 10.1007/978-81-322-1629-2__8, © Springer India 2014

141

Page 155: Sustainability of Indian Microfinance Institutions ||

142 8 Summary of Findings, Implications and Conclusion

Thereafter the second phase, which is an intermediate participant selection phase,was undertaken to fulfil the third objective of this study. The third objective was toidentify a set of efficient and sustainable MFIs in India, whose managerial strategiescan be referred to or emulated by the other MFIs operating in Indian microfinanceindustry. Seven such efficient and sustainable peer MFIs were identified from asample of 50 Indian MFIs for the year 2009. For identifying the efficient IndianMFIs, a non-parametric technique called DEA was used. The sustainability of theefficient MFIs was assessed by formulating a Sustainability Diamond Model, whichcaptures four pertinent dimensions of an MFI’s sustainability—financial, risk,growth and social dimensions. The DEA model used in this study also undertook abenchmarking process for the sample MFIs, where by efficient and sustainable peerMFIs, which are reference groups to the other lesser efficient MFIs, was identified.The analysis depicted the extent of input minimization to be achieved by each of thelesser efficient MFIs in order to optimize their performance. Having achieved thethird objective of identifying the efficient and sustainable Indian MFIs, the studymoved on its final qualitative phase. This phase of the study begins by undertakinginterviews with the MFI managers of four out of the seven MFIs identified inthe penultimate phase of the study. These interviews were conducted with theintention of understanding how these MFIs are managing the five factors affectingand discriminating their OSS status. Confirmation was also sought from the MFImanagers about the relationships shared by these factors with OSS ratio, as observedin the quantitative phase of the study. The strategies used by each of these MFIs tomanage these five factors—portfolio risk factor, growth factor, institutional factor,development factor and cost-efficiency factor—were documented for the referenceof lesser efficient and unsustainable MFIs. The policy suggestions made by themanagers to facilitate the management of the five factors without exploitation ofthe poor were also discussed. The validity of these strategies and policy suggestionswere checked using available supporting quantitative figures associated with eachof these factors. Thereafter, to complete the qualitative phase, a literature-baseddiscussion on the issues related to sustainability mismanagement was presented,taking the crisis that hit the industry in the year 2010 as the basis. Thus together thethree analytical phases pursued in this study explain the diverse issues associatedwith sustainability of Indian MFIs. Overall through this research investigationspanning across three phases, the author attempts to set the priorities right forsustainability issues in Indian MFIs and thereby aims to revive the lost hope inthe sector. The summary of findings drawn from each of these three phases and itsimplications are discussed in this concluding chapter.

8.2 Summary of Findings

The findings from the three analytical phases undertaken in this study are enumer-ated in the following subsections:

Page 156: Sustainability of Indian Microfinance Institutions ||

8.2 Summary of Findings 143

8.2.1 Findings from Phase 1: Quantitative Phase

In this phase, using multiple regression and multiple discriminant analysis tech-niques, the factors that affect and discriminate the OSS status of Indian MFIs wereidentified. The findings of this quantitative phase are summarized below:

(i) Portfolio risk factor (portfolio at risk greater than 30 days), growth factor (grossloan portfolio), institutional factor (credit delivery model) and developmentfactor (average loan size per borrower) were found to be the significant factorsaffecting OSS of Indian MFIs.

(ii) Growth factor was found to have a positive influence on the OSS of IndianMFIs. Portfolio at risk greater than 30 days, a portfolio risk factor, SHG creditdelivery model, an institutional factor and Average Loan Size Per Borrower, adevelopment factor, were found to share a negative influence on OSS of IndianMFIs.

(iii) Cost-efficiency factor (operating cost per borrower) was found to be the singlesignificant discriminant factor with the power to predict the OSS status ofIndian MFIs.

8.2.2 Findings from Phase 2: Intermediate ParticipantSelection Phase

In this intermediate participant selection phase, by framing a DEA model andSustainability Diamond Model, the efficient and sustainable peer Indian MFIs wereidentified. The efficiency analysis, benchmarking and sustainability assessmentundertaken in this phase resulted in the following findings:

(i) Seven efficient and sustainable peer Indian MFIs, which charge a reasonableinterest rate as set by the regulator or lower, were identified in this phase. Theseseven MFIs serve as peers and their managerial strategies can be referred to oremulated by the other MFIs operating in Indian microfinance industry.

(ii) The average pure technical efficiency score for the sample MFIs was found tobe 87.5 %. This depicted that the sample MFIs shall optimize its operationsby decreasing 12.5 % of its inputs without affecting its existing output levels.Similarly, the average scale efficiency score of the sample was found to be94.8 %, indicating a difference of 5.2 % between actual scale of operation andthe optimal scale of operations for the sample MFIs.

(iii) The returns to scale at which the sample MFIs are operating were identified.The maximum productive scale size at which the efficient and sustainable MFIsare operating was also presented for the reference of other MFIs.

(iv) The extent of input minimization that other MFIs in the sample should achieveto attain efficiency in their operations was ascertained in this phase.

Page 157: Sustainability of Indian Microfinance Institutions ||

144 8 Summary of Findings, Implications and Conclusion

8.2.3 Findings from Phase 3: Qualitative Phase

In this phase, the strategies used by four efficient and sustainable Indian MFIsto manage the five factors identified in the quantitative phase—portfolio riskfactor, growth factor, institutional factor, development factor and cost-efficiencyfactor—were identified and documented for the reference of other Indian MFIs.The relationships shared by these factors with the OSS ratio, as observed in thequantitative phase of the study was explained by the managers, by mapping it on tothe OSS ratio. The issues involved in the mismanagement of these factors are alsoportrayed, to pre-empt scenarios where the MFI loses its focus on social goals, inits pursuit for sustainability. The explanation on the strategies and relationships aresummarized as follows:

(i) Portfolio Risk Factor: The MFIs cited portfolio risk as a crucial sustenancefactor that denotes the recovery performance of their MFIs. The negativerelationship shared by this factor with OSS ratio was attributed by the managersto a positive correlation that exists between portfolio risk and loan lossprovisions. This positive correlation when tested on the sample data of 50Indian MFIs was found significant and therefore served as a validation for themanager’s observation. It was observed that this factor is best managed byadopting risk preventive strategies that emphasize on training of MFI creditofficers, provision of insurance coverage and capacity building services toclients. Risk detective and corrective strategies that emphasize on followingup of delinquent loans and monitoring of risk parameters (portfolio at riskgreater than 30 days and write-off ratio) were also recommended as a meansto ensure portfolio quality. But the managers considered risk prevention to beof more importance for an MFI, as recovery of delinquent loans was oftenvery difficult in their uncollateralized operating model. The MFIs mostly hadto meet the bad loans from the loan loss provisions, if delinquency occurred.Though defaults need to be curtailed, attaining an impressive portfolio qualityand aiming for frantic growth plans, by increasing the caseload of creditofficers with their incentives tied to zero delinquency rate targets, can result inover-indebtedness of clients and coercive loan recovery from them. Thereforethough portfolio risk management is pertinent for sustainability, the discus-sions presented in Chap. 7 show that care needs to taken from the managementside to avoid these indiscriminate acts that goes against the social goals ofthe MFI.

(ii) Growth Factor: Growth factor was found to be managed by growth-enhancingstrategies that result in vertical and horizontal expansion. The choice betweenvertical and horizontal expansion plans of the MFI was found to be made inrelation to the capital base available to support the additional costs involvedin horizontal growth. Vertical growth was achieved by standardized productdelivery and market penetration in familiar demographies. On the other hand,horizontal growth was achieved by replicating successful branch models innew markets. Customized products delivery was undertaken in the new market.

Page 158: Sustainability of Indian Microfinance Institutions ||

8.2 Summary of Findings 145

This required training of newly hired staff in these areas by the existing staff ofthe MFI. The cost and risks associated with the new operations required closemonitoring which was best done using a MIS. In addition to these growth-enhancing strategies, since the MFIs have experienced a positive relationshipbetween credit officer productivity and portfolio risk, they also advocated theuse of growth balancing strategies that maintains outreach at acceptable levelsof portfolio risk. MFI managers observed that frantic growth strategies, withoutconsidering its impact on the cost and risk of MFIs, would not contributeto the sustenance of the MFI in the long run. They observed growth to becontributing to the sustainability of the MFIs on two fronts—scale economiesand revenue enhancement. The sample data of 50 Indian MFIs corroborated thestrong positive contribution that growth makes towards revenue enhancementbut showed no evidence of significant scale economies on this account. InChap. 7, the discussions depict that, with this revenue enhancement potentialattached to growth, there are chances for management to lose focus on socialimpact and succumb to the demands of investors who fuel expansion plans withtheir capital. Investor pressures for achieving returns maximization may leadto levy of exorbitant interest rates, multiple lending and coercive loan recoverypractices, all of which goes against the spirit of this sector. It is recommendedthat MFI managers should take care to ensure that they do not get carried awayby investor pressures, for scaling up would have no meaning for MFIs, if itresults in an adverse impact for clients. Choosing social inclined investors tosupport growth is critical here.

(iii) Institutional Factor: In the case of the institutional factor (credit deliverymodel), the MFI managers observed that the choice between Grameen modeland SHG model can be done based on the model’s cost-effectiveness andapplicability to client’s empowerment needs. All the managers interviewedobserved that the operating cost associated with the Grameen model is lesswhen compared to the SHG model. This observation was validated by t testsconducted on the data of 50 Indian MFIs. The results confirmed that cost-wiseGrameen model was found to be more economical than SHG model, withno significant difference in portfolio risk and yield. The MFIs attributed thiscost advantage to the lesser group formation cost associated with the formermodel. Two of the efficient and sustainable MFI managers used this cost-effectiveness criteria for selection of the model, and therefore they preferredthe Grameen model to the SHG model. But despite acknowledging this costadvantage, the two other MFI managers had a specific reason for using theSHG model. According to them the SHG model is more appropriate for dealingwith clients for whom credit is not the only missing link to development.The SHG model spends more time on nurturing the group and empoweringthe clients than the Grameen model, thereby equipping the clients to indulgein sustainable income-generating activities. Moreover, they observe that it ispossible to reduce the higher group formation costs of SHGs by enteringinto NGO-MFI partnerships. So in the view of these managers, with the useof NGO-MF I partnership strategy, both the benefits of cost advantage and

Page 159: Sustainability of Indian Microfinance Institutions ||

146 8 Summary of Findings, Implications and Conclusion

client empowerment can be combined in the SHG model. In addition to thisobservation, in Chap. 7, it was discussed that a demand-driven credit modelthat has clients at the centre of its operations would best ensure client welfarein microfinance.

(iv) Development Factor: All the managers interviewed negated the theoreticalbelief that development for the poor can be attained merely by providing smallaverage loan size per borrower. They observed that they are unable to managethe development factor, by providing small average loan size per borrower, asthe client’s financial needs remain unsatisfied at this level. According to themanagers, true development orientation lies in progressively providing largerloans size that meet their growing financial needs and repayment capacity,and therefore they advocate this practice. The validity of this practice wasnot checked by surveying the clients view on this aspect. As the study ismore institution oriented, we checked the validity of this strategy by collectingdata on the client retention rate of the MFIs. Since the client retention ratewas found to be higher for MFIs with larger average loan size per borrower,it was taken as an evidence for the higher client-satisfaction level at thislevel. Moreover, the discussion in Chap. 7 on mismanagement also provesthe onset of an emerging literature citing high loan size to be more beneficialfor client needs, as it prevents them from availing multiple loans. Accordingto the MFI managers, the average loan size per borrower in itself is a poorproxy for capturing mission drift and development orientation of an MFI’soperations. Therefore, they do not attribute the negative relationship observedin the quantitative phase of the study between average loan size per borrowerand OSS ratio, to indicate absence of mission drift in the MFI’s operations.They explain this negative association to be on account of the higher screeningand monitoring cost associated with larger-sized loans. This explanation wasvalidated by the presence of a significant positive correlation between operatingcost per borrower and average loan size per borrower, on the sample data of the50 Indian MFIs.

(v) Cost-Efficiency Factor: Cost reduction strategies that reduce the operatingcosts and financing costs of the MFIs were discussed. Reduction of operatingcost was regarded pertinent, as it was found to be the major discriminatingaspect between an operationally self-sustainable and unsustainable Indian MFI.Since operating cost per borrower is a variable component of an MFI’s OSSratio, when compared to the other components of the ratio (other components,namely, revenue, financing cost and loan loss provisions comes within afixed range for most MFIs), the managers cited the former to play a crucialdiscriminatory role in determining he OSS status of an MFI. The efficientand sustainable MFIs were found to have reduced their operating costs mainlyby increasing credit officer productivity, through NGO-MFI partnerships, byusage of IT-enabled MIS and by adopting cash flow management techniques.Though the strategy of enhancing credit officer productivity resulted in costreduction for the MFI, it was also cited to have resulted in downside risk forthe MFI, by increasing its portfolio risk. Regression analysis undertaken on

Page 160: Sustainability of Indian Microfinance Institutions ||

8.3 Implications of the Findings 147

the sample data confirmed this observation and therefore it is recommendedthat the managers be conscious of the risk aspect while using this strategy.The discussions of mismanagement of cost factor also reiterate the probabilityof the same risk. The usage of an IT-enabled MIS was cited as a means toreduce operational inefficiencies and is recommended as a means to monitorcost, growth and risk parameters concomitantly. Entering into NGO-MFI part-nerships was cited as an effective means to reduce the group formation costsin a SHG model. The role of cash management in ensuring effective utilizationof cash flows was also emphasized as a means to reduce operating costs. In thecase of cost of finance, MFI managers observed that it is almost uncontrollableas cost of funds always averages around 12–13 % for majority of MFIs inthe industry. But MFIs with low operating costs, observed that negotiatingwith donors and investors to lend funds at a lesser rate has been successfulfor them, owing to their reputation of being an operationally efficiency MFI.MFIs also suggested the adoption of securitization and portfolio buy-out routesas a means for refinancing their loans. These practices have enabled them toinject large amount of capital at a price lower than the average cost of fundsfor Indian MFIs.

8.3 Implications of the Findings

The implications from the findings of this research work are categorized insubsections as 8.3.1, 8.3.2, and 8.3.3.

8.3.1 Practical Implications

The findings of this study have the following implications for MFI practitioners inIndia:

(i) The quantitative phase of the study identifies the determinants and discrim-inants of Indian MFI’s OSS status and recommends Indian MFI managersto concentrate on management of these factors to enhance their MFI’s OSSstatus. The identified factors requiring managerial attention are portfoliorisk factor, growth factor, institutional factor, development factor and cost-efficiency factor.

(ii) The study coins an OSS Predictor Model, which identifies cost-efficiencyfactor as the predominant discriminator of the sustainability of Indian MFIs.It recommends Indian MFI managers to use this model to predict their MFI’sOSS status 2 years from the date of estimation.

(iii) The efficiency analysis undertaken in the study enables Indian MFI managersto understand their MFI’s relative efficiency performance in the industry. The

Page 161: Sustainability of Indian Microfinance Institutions ||

148 8 Summary of Findings, Implications and Conclusion

extent of inefficiencies to be trimmed-off from the operating structure of thesample MFIs are depicted in the study. Thus, by knowing how well they areutilizing their resources and where to minimize inputs in their operations toimprove their current performance, the MFIs can enhance their efficiency.

(iv) The sustainability assessment model formulated in this study can assess thesustainability of MFIs from multiple dimensions—financial, social, risk andgrowth dimensions. Indian MFI rating agencies are recommended to use thisSustainability Diamond Model as it can enable them to rate MFIs based onthese multiple dimensions. As the model regards interest rates levied by MFIsas a social performance indicator, it enables in the identification of IndianMFIs which pass on the benefits of their operational efficiency to clients inthe form of reasonable interest rate.

(v) The study benchmarks the sample MFIs, so that the Indian MFIs can identifythe efficient and sustainable peer MFIs, whose strategies they can refer to oremulate to enhance their efficiency and sustainability status.

(vi) It documents the managerial strategies used by the efficient and sustainableIndian MFIs to manage the determinants and discriminants of their OSS. Adiscussion on the mismanagement of the factors probable if the MFIs loseits focus on client welfare is also presented taking crisis as the reference.This is expected to serve as a guide and a valuable learning experience forMFIs operating in Indian microfinance industry in terms of understanding thedifferent aspect of sustainability management.

(vii) In the qualitative phase of the study, the managers explain the portfolio riskfactor to share a positive relationship with loan loss of an MFI. They observethat once loans become delinquent, it is almost out of the MFI’s controland recovery becomes a formidable task. Therefore, they discuss portfoliorisk factor to be best managed by preventive strategies that emphasize ontraining of MFI credit officers, provision of insurance coverage and capacitybuilding services to clients. Indian MFI managers are therefore recommendedto adopt such preventive strategies to pre-empt loans from being delinquent,rather than incentive zero default rates and thereby inculcate the tendency ofcoercive recovery practices in the sector.

(viii) In the qualitative phase of the study, it was found that though growth factorpositively contributes to sustainability by enhancing operational income ithas not resulted in significant scale economies for the sample MFIs. Theefficiency analysis also observes a 5.2 % difference between actual andoptimal scale of operations for the sample MFIs. It depicts 28 % of the lesserefficient MFIs to be operating at decreasing returns to scale and 46 % to beexperiencing increasing returns to scale. Therefore, it is recommended thatthese lesser efficient MFIs optimize their scale of operations by referring tothe most productive scale size at which their efficient and sustainable peersare operating. However scaling up needs to be gradual and steady increasingthe impact for clientele, without resulting in strained borrower to credit officerrelationships and succumbing to investor pressures.

Page 162: Sustainability of Indian Microfinance Institutions ||

8.3 Implications of the Findings 149

8.3.2 Theoretical Implications

The findings of this study contribute to microfinance performance managementliterature. The efficiency analysis undertaken in this study differs from the existingstudy literature in that it formulates a DEA model that takes into account thedual goals of Indian MFIs, benchmarks the Indian MFIs based on their relativeefficiency scores, frames a sustainability assessment model and arrives at a list ofefficient and sustainable Indian MFIs. By denoting the extent of technical and scaleinefficiencies in the sample, the analysis proves that though Indian MFIs are globallyrenowned as least cost players, vast majority of them still have scope for trimming-off inefficiencies from its operating structure.

The sustainability analysis undertaken in this study also contributes to literature.Prior sustainability performance assessment undertaken in Indian context inferredthat sustainability can be enhanced by charging increased cost-covering interest rate,increasing credit officer productivity, retaining average loan size at low levels (nomission drift) and using SHG model. The findings of this study add to literature onall these fronts, in the following ways:

(i) On the Practice of Charging Cost-Covering Interest Rates: With respect tothe practice of levying cost-covering interest rate, this study is of the viewthat the operating costs of the MFIs should not be loaded with operationalinefficiencies of the MFIs. Therefore, the study advocates that MFIs shouldstrive to attain efficiency in its operations, before charging a cost-coveringinterest rate. Moreover, since interest rate cap is imposed on Indian MFIs,enhancing sustainability by increasing interest rates is no longer a validpractice. Therefore, the study identifies operationally efficient MFIs, whichremain sustainable by levying a reasonable interest rate as set by the regulatoror even lower and understands how such MFIs are managing their OSS.This is expected to exhort more and more MFIs to shed off their operationalinefficiencies and attain sustainability. The issues of mismanagement that canlead an MFI to lose its client focus, in the pursuit of sustainability, are alsodiscussed so as to pre-empt such happenings in the sector.

(ii) On the Practice of Enhancing Credit Officer Productivity: In tune with theexisting literature, the findings of this study indicate an inverse relationshipbetween credit officer productivity and operating costs of MFIs. Along withthis cost advantage, the study also observes a concomitant increase in theportfolio riskiness of the MFIs. Thus the study observes the existence of adownside risk associated with this cost reduction strategy that is vulnerable tocompromise on the portfolio quality of an MFI. Though the associated riskincrease is small in magnitude, there exists a significant positive relationshipbetween credit officer productivity and portfolio riskiness of MFIs when thecaseload of credit officers increases beyond 550 borrowers. Since recoveryperformance is critical for MFI’s sustainability, understanding this relationshipis pertinent for an MFI’s sustainability management. The discussion on

Page 163: Sustainability of Indian Microfinance Institutions ||

150 8 Summary of Findings, Implications and Conclusion

mismanagement of sustenance factors also portrays how increasing creditofficer caseload can hamper relationships on field with the clients, leading tocases of coercive and overbearing recovery practices.

(iii) On the Mission Drift Issue: Akin to the findings in existence literature,the results of the quantitative phase of this study indicate no evidence oftrade-off between sustainability and depth of outreach in Indian microfinanceindustry. Though in literature this an indication of absence of mission drift, themanagers interviewed in the qualitative phase explain this negative relationshipshared by loan size with OSS to be on account of the higher screening andmonitoring costs associated with larger loan size. They negate the theoreticalbelief predominant in microfinance literature, which attribute this negativerelationship to signify absence of mission drift. They observe that provision ofsmall average loan size per borrower does not connote development orientationin true sense, because by limiting their loan size, an MFI denies the fulfilmentof the growing financial needs of the poor. By partially fulfilling client needs,the MFIs do not retain its customers and faces the risk of losing their customersto other informal players, defeating its own mission. This was seen to result inmultiple-borrowing tendency among the customers. Thus, the study unveils thebeginning of eclectic views that exists among practitioners and theorists, aboutthe usage of average loan size per borrower, as a proxy for an MFI’s adherenceto its development mission of serving the poor.

(iv) On the Usage of Self-Help Group Credit Delivery Model: Contrary to theexisting literature, this study finds the SHG model to be less cost-effective thanthe Grameen model, with no significant difference in portfolio risk and yieldacross these models. The prior study in literature which concluded SHG modelto be cost-effective was undertaken for the year 2003, when vast majority ofthe Indian MFIs used the home-grown SHG model. But since the period ofthis study (2004–2009) had a more representation for Grameen replicators,comparison of costs between the two models was possible. The higher costsfor SHG model was attributed by the MFI managers to the higher groupformation and nurturing costs involved in it. Though SHG was found tohave higher operating costs, MFIs using this model had a reason for usingit. Owing to the fact that the SHG model nurtures the client groups for alonger period, its usage was suggested for those clients who needed suchsocial assistance in addition to financial access. Moreover, it was found that byentering into NGO-MFI partnerships, the group formation cost in SHG modelcan be reduced considerably, combining the benefits of cost advantage andclient empowerment in the SHG model. Thus the findings of this study addto literature by depicting that both SHG model and Grameen model are beingused by efficient and sustainable MFIs, based on the model’s cost-effectivenessand applicability to client’s empowerment needs. But the discussion on thecrisis shows that the more a model is designed to be demand-driven, makingclients at the core, by the usage of democratic processes, the better would bethe benefits that it can confer to the clients.

Page 164: Sustainability of Indian Microfinance Institutions ||

8.3 Implications of the Findings 151

8.3.3 Policy Implications

The findings of the study have policy implications, with recommendations to theIndian MFI regulator, RBI. The recommendations are as follows:

(i) The Malegam Committee, which was specially constituted to study the post-crisis issues of Indian microfinance industry, observed 26 % interest rate to bea reasonable pricing for Indian MFIs. It mandates all Indian MFIs to ceil theirinterest rate at this reasonable interest rate. This raised huge concerns amongMFI practitioners as to whether this interest rate ceiling is high enough forMFIs to flourish and remain sustainable. This study finds that efficient MFIscan remain sustainable, at 26 % and even below this rate. Efficient IndianMFIs like Sanghamithra, SKDRDP, Sarvodaya Nano Finance and Pustikarare able to remain sustainable by charging interest rate in the range of 13–16 %. It was found that age of MFI, size of MFI, credit model, regulatorystatus or area of operation had no significant influence on the efficient andsustainable status of these MFIs. Efficient and sustainable MFIs attribute thissolely to the managerial strategies used by them. Therefore, it is suggested thatpolicymakers take active interest to disseminate the strategies used by efficientand sustainable MFIs to the rest of players in the industry, along with the areaswhere there are scope for mismanagement. This in turn would facilitate moreand more Indian MFIs to focus on client welfare and enhance their efficiencyand sustainability, thereby resulting in the discovery of a fair and reasonableinterest rate lower than present ceiling.

(ii) The study finds that by the usage of an MIS, the efficient and sustainableIndian MFIs are able to get an integrated view of the determinants anddiscriminants of their OSS status on a perpetual basis. This transparency inoperations enables information-based management at the MFI level and alsofacilitates compliance with the information reporting requirements of the RBI.Therefore, it is recommended that RBI insist the installation of a MIS asa mandatory requirement for all Indian MFIs licensed as NBFC MFIs. Itis also recommended that RBI insist NGO-MFIs desirous of transformingto a regulated status of NBFCs to install a MIS 1 year prior to seekingregistration. Compliance with all the information requirements of a NBFC-MFI must be made mandatory during this year and licence as a regulatedNBFC shall be granted subject to the adherence to this norm. Organizationslike Small Industries Development Bank of India (SIDBI) and NABARDare recommended to support the endeavours made by MFIs to implementIT-enabled MIs, by providing technology specific loans and grants to them.Assistance can also be provided to MFIs through the Microfinance EquityFund, proposed by the Finance Minister of India to support small MFIs. Theusage of an MIS can facilitate transparency in microfinance operations andovercome the data constraints that currently prevail in the Indian microfinanceindustry. With this enhanced transparency, the efficiency of MFI operationscan be monitored by regulators. It can encourage information sharing in theindustry which in due course can aid more MFIs to be efficient.

Page 165: Sustainability of Indian Microfinance Institutions ||

152 8 Summary of Findings, Implications and Conclusion

(iii) Apart from these suggestions emanating from the findings of the study, duringthe expert interviews, the MFI managers have also put forth some suggestionsto the regulator to facilitate the management of the sustenance factors ofMFIs, without exploiting the poor. Formation of a credit information bureau,complaint redressal mechanism, removal of caps of average loan size perborrower, relaxation of norms for deposit mobilization and recognition of cost-efficient MFIs are the suggestions made by MFIs to regulators, which warrantsdue regulatory attention.

8.4 Limitations of the Study

The limitations of the three phases of this research work are discussed in thefollowing subsections:

8.4.1 Limitations of the Quantitative Phase

(i) Owing to data constraints a longitudinal study on a larger sample size couldnot be undertaken.

(ii) Though governance was a variable depicted in literature to have an influenceon the sustainability of MFIs, it could not be captured separately in this study,as there was no comparable data available on this aspect for the entire sampleof MFIs.

(iii) A parsimonious model that takes care of the micro-level factors affectingOSS of an MFI is undertaken in this study. Therefore, the influence ofmacroeconomic factors on the OSS status of MFIs is beyond the purview ofthis study.

8.4.2 Limitations of the Intermediate Participant SelectionPhase

(i) The concept of efficiency used in this study is relative in nature.(ii) A time series analysis of efficiency using Malmquist index could not be

undertaken as there was no continuous data available for all the input–outputvariables in the DEA model.

(iii) The models used to identify the efficient and sustainable MFIs have used input–output parameters which are internal to the MFI. As a result, the interactionswith the economy and the impact on client and societal expectations andperceptions are not reflected in the empirical model. In this sense the model

Page 166: Sustainability of Indian Microfinance Institutions ||

8.5 Scope for Further Work 153

has limitations in capturing the sustainable growth of the best performer peerMFIs on a long-term basis. Owing to this reason, the MFIs featured as efficientand sustainable in the study, concord more with short-term sustainabilitythan long-term sustainability. Long-term sustainability would call for viewingsustainability management issues, with reference to its impact on clientele andthis is captured partially using a literature-based approach, subsequently inChap. 7.

8.4.3 Limitations of the Qualitative Phase

(i) This phase experiences the inherent limitations of interview method. Thefindings of this phase are dependent on the skill of the interviewer in elicitingexplanation from the MFI managers and also on the level of participation andexpertise of the interviewee.

(ii) The strategies discussed by the MFI managers, though based on managerialexperience, will have some element of subjectivity in it. Considering thisaspect, earnest efforts were put in to triangulate and validate the strategiesusing available quantitative figures. But unavailability of perfect surrogates forvalidation was a limitation in this process.

(iii) The issues on microfinance crisis and mismanagement of sustenance factorscould have been backed with interviews but was limited to a literate-based dis-cussion; as after the out-break of the crisis, players were busy contemplating onthe pros and cons of the forthcoming regulatory changes, making respondentavailability a constraint.

8.5 Scope for Further Work

Subject to data availability in future, this work can be replicated on a larger samplesize to check if the significance of the factors still holds valid. A longitudinal studyin this regard, which captures the perceptions of different stakeholders in the postmicrofinance crisis period, can also be conducted. Based on the willingness ofthe MFIs to share data on their governance issues, investigations can be made tounderstand the role of governance on the sustainability of MFIs. The eclectic viewson the use of average loan size per borrower to denote the developmental mission ofan MFI can be further validated by research works that take into account the clientperspective on this issue. To further validate the strategies used by the efficient andsustainable MFIs and to test its applicability and chances of mismanagement, acrossdiverse operating conditions, an action-oriented research work can be undertaken.Moreover, the impact of the operations of the efficient and sustainable MFIs on itsclients and society and its interactions with economy at large can be studied in detailto assess the long-term sustainability of these institutions.

Page 167: Sustainability of Indian Microfinance Institutions ||

154 8 Summary of Findings, Implications and Conclusion

8.6 Conclusion

This book concludes with a note: Sustainability is a matter of pertinence for all MFIsto ensure perpetuity of its operations. Sustainability therefore deserves managerialattention as it is not something to be left to serendipity. Conscious efforts fromthe management side are required to make an MFI trim down its inefficienciesand make a move towards the path of sustenance. The study shows efficient andsustainable MFIs in India to have devised specific managerial strategies to ensurethat they turn a surplus out of their socially oriented operations. These strategiesare worth a reference for any other player aiming to trend the path of efficiencyand sustainability. The research investigation undertaken in this book depicts thatwith proper strategies in place, efficient MFIs are able to sustain operations, at thecapped rates or even much lower rates, ranging from 13 % to 16 %. But whilestriving for sustenance, as an unintended consequence of the best of the intentions,MFIs may experience the tendency to deviate its focus from the larger picture ofclient welfare. To portray these possibilities and to pre-empt its further occurrence,a discussion on sustainability mismanagement, with reference to the crisis that hitIndian microfinance sector, is brought forth in this book. The discussion depicts thatthough the enthusiasm for permanence, captured by the pursuit for sustainability,is commendable, it would lose all its sheen if the priorities are messed up in theprocess. Sustainability has significance for an MFI only as a means to the end ofachieving its social goals and not otherwise.

Page 168: Sustainability of Indian Microfinance Institutions ||

Appendices

Appendix 1: Operational Definition and Distinction BetweenOperational Self-Sustainability Ratio, FinancialSelf-Sustainability Ratio and Subsidy Dependence Index

Operating self-sufficiency (OSS) is a sustainability ratio which indicates whetheror not enough revenue has been earned to cover the MFI’s total costs—operationalexpenses, loan loss provisions and financial costs. A ratio above 100 % indicatesthat MFI has enough operating income to cover its costs, indicating an operationallyself-sustainable status.

OSS Ratio D Operating income .Loans C Investments/Operating costs C Loan loss provisions C Financing costs

Financial self-sufficiency (FSS) is a sustainability ratio that allows determinationof the extent to which operations of an MFI are becoming (increasingly) self-sustaining. Financial self-sufficiency indicates whether or not enough revenue hasbeen earned to cover both direct costs—including financing costs, provisions forloan losses and operating expenses—and indirect costs, including the adjusted costof capital.

The adjusted cost of capital is considered to be the cost of maintaining the valueof the equity relative to inflation (or the market rate of equity) and the cost ofaccessing commercial rate liabilities rather than concessional loans.

Adjusted cost of capital D [Inflation rate � (Average equity � Average fixedassets)] C [(Average funding liabilities � Market rate of debt) � Actual financingcosts]

FSS ratio D Operating income .Loans C Investments/

Operating costs C Loan loss provisions C Financing costsC Adjusted cost of capital

N. Marakkath, Sustainability of Indian Microfinance Institutions:A Mixed Methods Approach, India Studies in Business and Economics,DOI 10.1007/978-81-322-1629-2, © Springer India 2014

155

Page 169: Sustainability of Indian Microfinance Institutions ||

156 Appendices

Unless 100 % financial self-sufficiency is reached, the long-term provision ofcredit services will ultimately be undermined by the impact of inflation and thecontinued necessity to rely on donor funds.

The Subsidy Dependence Index (SDI) is an often less used measure of sus-tainability, though it is one of the best indicators of adjusted profitability from atechnical standpoint.

It measures how much an MFI would have to increase its lending interest rate tocover all of its costs including adjustments.

An SDI above zero means that the MFI still needs subsidy to operate—i.e. ithas not achieved financial sustainability. A two-stage calculation produces first theamount of annual subsidy and then the index.

S D A .m � c/ C ��E�m

� � P� C K (A.1)

where

S D Annual subsidy received by the MFIA D MFI concessional borrowed funds outstanding (annual average)m D Interest rate the MFI would be assumed to pay for borrowed funds if access to

borrowed concessional funds were eliminatedc D Weighted average annual concessional rate of interest actually paid by the MFI

on its average annual concessional borrowed funds outstandingE D Average annual equityP D Reported annual before-tax profit (adjusted, when necessary, for loan loss

provisions, inflation, and so on)K D Sum of all other annual subsidies received by the MFI (such as partial or

complete coverage of the MFI’s operational costs by the state)

SDI D SLP�i

(A.2)

where

SDI D Index of subsidy dependence of MFIS D Annual subsidy received by the MFI (see above)LP D Average annual outstanding loan portfolio of the MFIi D Weighted average interest yield earned on the MFI’s loan portfolio

Appendix 2: Charnes, Cooper and Rhodes Model and Banker,Charnes and Cooper Model Formulation

In 1978, Charnes, Cooper and Rhodes (CCR) formulated the CCR DEA modelas a fractional programming, which can be transformed to linear programming asfollows:

Page 170: Sustainability of Indian Microfinance Institutions ||

Appendices 157

Min™;œ™

s:t: � yi C Y: œ � 0;

™: xi � X: œ � 0; œ � 0 (A.3)

where X and Y are the K � N input matrix and the M � N output matrix (for the ithfirm, these are represented by the vector xi and yi), respectively. œ is a N � 1 vectorof constant and ™ is a scalar, which stands for efficiency of ith firm. By solving thislinear programming model for each of the N firms, the efficiency scores for each ofthe firms can be obtained. Model (1) is an input orientation DEA model under theassumption of constant returns to scale (CRS).

The CCR model assumes CRS and presupposes that there is no significantrelationship between the scale of operations and efficiency. But the CRS assumptionis valid only when all firms are operating at an optimal scale. Since in reality firmsexperience economies or diseconomies of scale, the overall technical efficiencyscores that are derived from this model are contaminated with scale efficiencies.Considering this limitation to account, the Banker, Charnes and Cooper (BCC)model was formulated in the year 1984.

BCC model relaxed the restriction of CRS to account for variable returns toscale (VRS) technology by adding convexity constraint to model (A.3). The VRSassumption provides the measurement of pure technical efficiency (PTE), which isthe measurement of technical efficiency devoid of the scale efficiency effects. TheBCC input orientation DEA model is as follows:

Min™;œ™

s:t: � yi C Y: œ � 0:

™: xi � X: œ � 0:

Xœ D 1; œ � 0 (A.4)

Model (A.3) and model (A.4) can be transformed to output orientation DEAforms as shown in model (A.5) and (A.6), respectively.

Maxø;œ¿

s:t: � ¿:yi C Y: œ � 0:

xi � X: œ � 0:

œ � 0 (A.5)

Page 171: Sustainability of Indian Microfinance Institutions ||

158 Appendices

Maxø;œ¿

s:t: � ¿:yi C Y: œ � 0

xi � X: œ � 0

Xœ D 1; œ � 0 (A.6)

Where, Y, X, xi, yi and œ are defined as previously in model (A.3); ¿ denotesproportional increase in output, which ranges from one to infinity.

Appendix 3: Interview Guide

1. Is the efficiency and sustainability status of your MFI dependent on any peculiarcharacteristic of your MFI—like age, credit delivery model, regulatory status,area of operations and size of MFI—or any other factors you consider relevant?

2. In the quantitative phase of the study, portfolio risk factor (portfolio at riskgreater than 30 days ratio) was found to share a negative relationship withoperational self-sustainability (OSS) ratio of Indian MFIs. Do you confirm theexistence of such negative relationship between portfolio risk factor (portfolioat risk greater than 30 days ratio) and OSS ratio of Indian MFIs? If yes, howwould you explain this negative association by mapping it to the OSS ratio? Ifnot, how would you explain the alternate relationship?

Operational Definitions:

Portfolio risk factor D Portfolio at risk greater than 30 days

Gross loan portfolio

OSS ratio D Operating income

.Operating costs C Financing costs C Loan loss provisions/

3. How do you manage portfolio risk in your MFI?4. In the quantitative phase of the study, growth factor (gross loan portfolio)

was found to share a positive relationship with operational self-sustainability(OSS) ratio of Indian MFIs. Do you confirm the existence of such a positiverelationship between growth factor (gross loan portfolio) and OSS ratio ofIndian MFIs? If yes, how would you explain this positive association bymapping it to the OSS ratio? If not, how would you explain the alternaterelationship?

Page 172: Sustainability of Indian Microfinance Institutions ||

Appendices 159

Operational Definitions:

Growth factor D Gross loan portfolio of the MFI

OSS ratio D Operating income

.Operating costs C Financing costs C Loan loss provisions/

5. How do you manage growth in portfolio operations in your MFI?6. In the quantitative phase of the study, development factor (average loan size

per borrower) was found to share a negative relationship with operational self-sustainability (OSS) ratio of Indian MFIs. Do you confirm the existence ofsuch a negative relationship between development factor (average loan size perborrower) and OSS ratio of Indian MFIs? If yes, how would you explain thisnegative association by mapping it to the OSS ratio? If not, how would youexplain the alternate relationship?

Operational Definitions:

Development factor D Average loan size per borrower of the MFI

OSS ratio D Operating income

.Operating costs C Financing costs C Loan loss provisions/

7. How do you manage development in your MFI through average loan size perborrower?

8. In the quantitative phase of the study, institutional factor (Self-Help GroupCredit Delivery Model) was found to share a negative relationship withoperational self-sustainability (OSS) ratio of Indian MFIs. Do you confirm theexistence of such a negative relationship between institutional factor (Self-HelpGroup Credit Delivery Model) and OSS ratio of Indian MFIs? If yes, how wouldyou explain this negative association by mapping it to the OSS ratio? If not, howwould you explain the alternate relationship?

Operational Definitions:

Institutional factor D Self � Help Group Credit Delivery Model

OSS ratio D Operating income

.Operating costs C Financing costs C Loan loss provisions/

9. How do you manage credit delivery model to enhance OSS of your MFI?10. In the quantitative phase of the study, cost-efficiency factor (operating cost per

borrower) was found to be a discriminating factor for the operational self-sustainability (OSS) status of Indian MFIs. Do you confirm the existence ofa discriminatory relationship between cost-efficiency factor (operating cost per

Page 173: Sustainability of Indian Microfinance Institutions ||

160 Appendices

borrower) and OSS status of Indian MFIs? If yes, how would you explain thisrelationship by mapping it to the OSS ratio? If not, how would you explain thealternate relationship?

Operational Definitions:

Cost-efficiency factor D Operating cost per borrower of the MFI

OSS ratio D Operating income

.Operating costs C Financing costs C Loan loss provisioning/

11. How do you manage operating costs in your MFI?12. How do you manage financing costs in your MFI?

Appendix 4: Summary of the Qualitative Data CollectedDuring the Interviews

Table A.1 Qualitative data collected from MFI A

MFI

MFI A

FactorsMapping to OSS ratio toexplain the relationship

Strategy used for managingthe factor Policy suggestion

Portfolio riskfactor

PAR > 30 days shares anegative relationshipwith OSS as it has apositive associationwith loan lossprovisions

Prevent vulnerability todefault by providingcapacity building andwelfare services toclients

Growthfactor

Gross loan portfolioshares a positiverelationship withOSS as it has apositive associationwith operatingincome

Balance growth withportfolio risk

To fix the problem ofmultiple lending,which is anadverse effect ofMFI’s growth, acredit informationbureau needs tobe place at theearliest

Increase of credit officerproductivity to achievegrowth must notcompromise portfolioquality

Awareness of the positiveassociation betweencredit officerproductivity andportfolio risk required

(continued)

Page 174: Sustainability of Indian Microfinance Institutions ||

Appendices 161

Table A.1 (continued)

MFI

MFI A

FactorsMapping to OSS ratio toexplain the relationship

Strategy used for managingthe factor Policy suggestion

Institutionalfactor

Though SHG creditmodel has highergroup formationcost, which can bereduced byNGO-MFIPartnerships

SHG is used as credit isnot the only missinglink to developmentfor our clients

High cost on SHGformation is reducedby outsourcing thistask to NGOs fornominal fees

Developmentfactor

Larger sized loan willhave higherscreening andmonitoring cost,which means higheroperating costs onthese loans

Progressively increasesloan size based onclient needs becausesmall loan size willnot always fulfil clientneeds

Cost-efficiencyfactor

Since operating costsaccount for nearly2/3 of an MFI’s cost,a lower operatingcost per borrower,can distinguish anMFI’s sustainabilitystatus

NGO-MFI partnershipsreduce groupformation cost.Similarly use of MISreduce operating costsand avoid the need formanual data entry.Increase of creditofficer productivity byrecruiting staff whoshare the samehousehold economicsof the clients are also ameans used forreducing operatingcosts. Financing costsare low, because thereputation of the MFIbeing a low costplayer, who chargeslow interest rate, helpsto negotiate withinvestors to lend fundsat low rate

The ceilings imposedon interest ratesand financialmargins, reducesthe incentive forattainingcost-efficiencyand levying lowinterest rate.Therefore theregulators shouldreward low costMFIs, whichcharge interestrate below theceiling, bygranting them thepermission tosource funds frombanks at belowbase rate

Page 175: Sustainability of Indian Microfinance Institutions ||

162 Appendices

Table A.2 Qualitative data collected from MFI B

MFI

MFI B

Factors

Mapping to OSSratio to explain therelationship

Strategy used for managing thefactor Policy suggestion

Portfolio riskfactor

PAR > 30 daysand loan lossprovisions aredirectly related

Providing welfare services toclients can enable them torun sustainablemicroenterprises, preventingany probable loan defaults

Formation of awell-functioningcomplaintredressal systemlike that of anombudsman tohandle clientcomplaints relatedto over-bearingand coerciverecovery practices

Following-up of defaultingloans and enforcement ofjoint-liability principle aremeans to addressdelinquency

Growthfactor

With higher grossloan portfolio,there will behigheroperatingincome

Achieve vertical expansionthrough market penetrationin existing markets.Achieved through renderingof standardized products andenhanced credit officerproductivity

Institutionalfactor

SHG model hashigher groupformation cost,but withNGO-MFIpartnershipswe reduce thiscost

SHG model is used because ourcustomers need not justfinancial inclusion but alsosocial inclusion

SHG model spends more timeon nurturing the group andempowering the clients thanthe Grameen model, therebyequipping the clients toindulge in sustainableincome-generating activities

The higher cost on groupformation gets reduced byentering into NGO-MFIpartnerships, to handle thegroup formation tasks

Developmentfactor

Larger sized loanshave higheroperating cost,because higherlevels ofmonitoringwill be neededon them

Small loan size will inducetendency among clients togo for multiple lending. Sowe progressively increaseloan size to meet thegrowing financial needs ofour clients

(continued)

Page 176: Sustainability of Indian Microfinance Institutions ||

Appendices 163

Table A.2 (continued)

MFI

MFI B

Factors

Mapping to OSSratio to explain therelationship

Strategy used for managing thefactor Policy suggestion

Cost-efficiencyfactor

Operating costbeing thelargestdenominator ofthe OSS ratio,a loweroperating costper borrowercandifferentiate itssustainabilitystatus

Enhancing credit officerproductivity is a means usedfor operating cost reduction.The use of back-end MISwith a personal digitalassistant, helps to reduceoperating cost by 5 %. Ithelps to gain a 360ı view ofthe five factors identified ascrucial for sustainability.Portfolio buy-out model isused to liquidate loans andsource finance at adiscounted rate from thebanks

Table A.3 Qualitative data collected from MFI C

MFI

MFI C

Factor

Mapping to OSS ratioto explain therelationship Strategy used for managing the factor

Policysuggestion

Portfolio riskfactor

When PAR > 30 daysdecreases, loanloss provisionexpenses alsoreduces

Prevent risk at early stage, by trainingcredit officers in surrogate creditassessment and customer-centricproduct development. This willensure smooth recoveries without useof any coercive practices. Monitorportfolio at risk >30 days ratio andwrite-off ratio. Ideally these riskindicators should be 10 % or below,to ensure control of risk withinacceptable parameters

Growthfactor

Gross loan portfolioshares a positiverelationship withOSS as it has anegativerelationship withoperating costs

Decide growth strategy afterunderstanding the costs involved inexpansion and the available capitalsupport

(continued)

Page 177: Sustainability of Indian Microfinance Institutions ||

164 Appendices

Table A.3 (continued)

MFI

MFI C

Factor

Mapping to OSS ratioto explain therelationship Strategy used for managing the factor

Policysuggestion

Institutionalfactor

SHG model hashigher operatingcost, as the groupformation cost ishigher in it than inthe Grameenmodel

Grameen model is used because it ismore institutional oriented, with lessoperating cost in group formation

Developmentfactor

It is morecumbersome tomonitor a largesized loan and thisresults in higheroperating costs

Providing small loan size alone does notmean an MFI is fulfilling itsdevelopmental mission of reachingthe poor. True development lies inbeing able to cater to the clients’increasing financial needs, withouthaving them resort to other informalsources

Cost-efficiencyfactor

Operating cost is amajor componentof the OSS ratio.So a loweroperating cost perborrower ratio,can differentiatesustainability

Ensuring efficient fund circulation, bykeeping the overnight cash at itsminimum is a means used forreducing operating costs. Enhancedcredit officer productivity and use ofMIS also helps to reduce operatingcost. Securitization of loans to banksis done for reducing cost of funds

Table A.4 Qualitative data collected from MFI D

MFI

MFI D

Factor

Mapping to OSSratio to explain therelationship

Strategy used for managingthe factor Policy suggestion

Portfolio riskfactor

There is a directrelationshipbetweenPAR > 30 daysand loanimpairmentexpenses

Train field officers to assessthe income of the clientsand design financialproducts in tune with thecash-flow patterns of theclients, to prevent risk

Provide insurance coverage toclients to prevent loss onaccount of uncontrollablefactors adversely affectingrepayment

(continued)

Page 178: Sustainability of Indian Microfinance Institutions ||

Appendices 165

Table A.4 (continued)

MFI

MFI D

Factor

Mapping to OSSratio to explain therelationship

Strategy used for managingthe factor Policy suggestion

Growthfactor

Gross loanportfolio sharesa positiverelationshipwith OSS as ithas a negativerelationshipwith total costsof MFI

Achieve horizontal growththrough replication ofsuccessful branch model

Achieved through providingcustomized products

Institutionalfactor

Group formationcost and time ishigher in SHGmodel than inthe Grameenmodel

Grameen model is usedbecause it forms groupsand immediately disbursesloans

As the group formation timeand cost is far less inGrameen compared toSHG model, the overalloperating costs are low

Developmentfactor

Larger sized loanswill requiremore screeningefforts and thiswill result inhigheroperating costs

Clients resort to multipleborrowings from informalsources and moneylendersfor meeting their growingnon-productive andnon-routine expenses. Thisis despite the fact that theinterest rates charged bythese sources are higherthan that levied by MFIs.This is so because theformer sources do notimpose any constraints onloan size

To remove caps imposedby regulator on MFIloan size, as it willinduce the tendencyfor multipleborrowings amongthe clients. Only aflexible loan size thatmatches therepayment capacityand financial needsof the clients willresult in truedevelopment, notmere provision ofsmall-sized loans

Cost-efficiencyfactor

Operating costbeing thelargest chunkof an MFI’scost structure,itssustainabilitywill bedifferentiatedon this basis

Operating costs are reduced byincreasing the borrower tocredit officer ratio and bythe use of a MIS. MIS havereduced operationalexpenses close to 5 %.Portfolio buy-out model isused for reducing cost offunds

The present regulationfor depositmobilization needsrelaxation to accountfor rating theuncollateralizedmodel in which anMFI operates. This isessential forenhancing depositmobilization, whichcan serve as sourceof capital for the MFI

Page 179: Sustainability of Indian Microfinance Institutions ||

166 Appendices

Appendix 5

Table A.5 Group formation process in 14 days

Phases Process

Village/slum selection phase (3 days) In case of vertical growth: select area within 20 km of anexisting branch office

In case of horizontal growth: select area where branchreplication is possible

Target areas with 240 households such that there isminimum of 60 client potential in a village &minimum of 3,000–4,000 client potential in a slum

Assess market potential based on secondary datacollected from government offices

Introductory phase (2 days) Open to all village/slum dwellersGive overview of MFI and loan products to potential

target clientsClient selection phase (3 days) Target economically active poor woman (in the age

group 18–55 years), i.e. either self-employed orwage labourer

Monthly income criteria: individual <Rs1,500;family <Rs2,500

Resident of village/slum for at least one preceding year,not owning a large house

Willing to be a guarantor for other women in the groupGroup formation phase (1 day) Out of the eligible clients, groups comprising of five

self-chosen members, who reside in the same villageare formed

Who share same socio-economic background and havemutual trust

No more than one member from the same household in agroup

Each group elects its own leader among the membersFour groups federate to form a centre, headed by centre

leaderGroup training phase (5 days) A 5-day training in the local language with easy,

numerical illustrations on products to ensure thatclients understand the MFI’s product, joint-liabilityconcept, terms and conditions and the creditdiscipline expected of them

Group recognition test to ensure that clients know andtrust their group members