67997 Access to Finance FORUM Reports by CGAP and Its Partners No. 8, February 2014 Greenfield MFIs in Sub-Saharan Africa A Business Model for Advancing Access to Finance Julie Earne, Tor Jansson, Antonique Koning, and Mark Flaming This research was funded by the Partnership for Financial Inclusion, a joint initiative by The MasterCard Foundation and IFC that aims to scale up commercial microfinance and advance mobile financial services in Sub-Saharan Africa. An important objective of the Partnership is to contribute to the global community of practice on financial inclusion, and to share research and lessons learned for the common good. The MasterCard Foundation advances microfinance and youth learning in developing countries to promote financial inclusion and prosperity Authors The authors are Julie Earne and Tor Jansson from IFC, Antonique Koning from CGAP, and Mark Flaming, independent consultant. Flaming served as IFC nominee on the board of MicroCred. After completing the research and writing for this publication he joined MicroCred as their COO. Acknowledgments This paper and the research behind it have been jointly produced by CGAP and IFC, with generous support from The MasterCard Foundation. IFC has direct investments in six of the holding companies and 17 greenfield MFIs that participated in the research. The authors are grateful to the many networks and MFIs who have contributed data and time in support of this project, including Access Holding, Accion, Advans Group, ASA, BRAC, EcoBank, FINCA, MicroCred Group, Opportunity International, ProCredit, and Swiss Microfinance Holding. We would also like to thank members of our esteemed advisory committee—Blaine Stephens, Michael Mithika, Rene Azokli, and Alexia Latortue—for being so generous with their time and advice and Greg Chen and Kate McKee at CGAP and Momina Aijazuddin, Alexis Diamond, and Greta Bull at IFC for their reviews and insightful comments. We also thank Anna Nunan at CGAP for her professional editing. Special thanks goes to Adam Sorensen of IFC and Hannah Siedek for their dedication and significant contributions in researching and organizing data central to this analysis and to Fahima Bille for supporting these efforts. Siedek is the former CEO of ProCredit DRC and assisted with interviews and market analysis for the DRC. The market analysis for the other two markets was done by Yaw Brantuo, an independent consultant, for Ghana, and a team of FTHM Conseils for Madagascar. © 2014, Consultative Group to Assist the Poor (CGAP) and International Finance Corporation (IFC) 1818 H Street NW, MSN P3-300, Washington DC 20433 Internet: www.cgap.org Email: cgap@worldbank.org Telephone: +1 202 473 9594 Rights and Permissions This work is available under the Creative Commons Attribution 3.0 Unported license (CC BY 3.0) http:// creativecommons. org/licenses/by/3.0. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Cite the work as follows: Earne, Julie, Tor Jansson, Antonique Koning, and Mark Flaming. 2014. “Greenfield MFIs in Sub-Saharan Africa: A Business Model for Advancing Access to Finance.” Forum. Washington, D.C.: CGAP and IFC. License: Creative Commons Attribution CC BY 3.0 Translations—If you create a translation of this work, add the following disclaimer along with the attribution: This translation was not created by CGAP and IFC and should not be considered an official translation. CGAP and IFC shall not be liable for any content or error in this translation. All queries on rights and licenses should be addressed to CGAP Publications, The World Bank Group, 1818 H Street NW, MSN P3-300, Washington, DC 20433, USA; e-mail: cgap@world bank.org. Contents Executive Summary 1 SECTION 1. Introduction 3 1.1 Landscape 3 1.2 Sample Cohort 5 SECTION 2. Greenfield MFI Performance 9 2.1 Balance Sheet Indicators 10 2.2 Growth and Operational Performance 12 2.3 Financial Performance 14 SECTION 3. The Holding Company 17 3.1 Structure 17 3.2 Investors 18 3.3 Mission and Strategy 21 3.4 Institutional Capacity and Knowledge Transfer 22 3.5 Success Factors, Common Challenges 23 SECTION 4. The Role of Greenfield MFIs in Market Development 25 4.1 Market Share 25 4.2 Skills Building 26 4.3 Product and Channel Diversification 27 4.4 Standards and Good Practices 28 SECTION 5. Conclusion 31 ANNEX 1 Methodology 33 ANNEX 2 List of People Interviewed 35 ANNEX 3 Bibliography 37 i ii Executive Summary S ub-Saharan Africa (SSA) has the lowest level of access to nance in SSA. A good number of greenfield MFIs now have a finance of any region in the world, with an average sufficient track record to enable an analysis of their perfor- banked population of only 24 percent (Findex 2012). mance and role in the market. This publication includes The region’s banking systems are small in both absolute and some references to other types of financial service providers relative size, and the microfinance sector has been relatively in SSA in a limited way and only to provide an overview of slow to expand in SSA compared to other regions in the world. the spectrum of providers, not to give a quantitative compar- There is a range of strategies for extending the reach of micro- ison between different models. This stocktaking of the finance, including the transformation of existing institutions, greenfield experience should help inform decisions for vari- the creation of stand-alone greenfield microfinance institu- ous stakeholders. Specifically the paper will inform the com- tions (MFIs) with and without a centralized management or ing generation of investment in microfinance, including how holding structure, bank downscaling, and others. much and what kind of funding is necessary to support ca- This Forum explores the contribution of greenfield MFIs pacity building in new institutions as well as those ready to to access to finance in the region. The greenfield business scale up. The paper may also help to promote regulatory con- model is focused on expanding financial services through two sistency for institutions providing a full menu of micro, small, main elements: (i ) the creation of a group of “greenfield MFIs” and medium enterprise services. defined as institutions that are newly created without pre-ex- Section 1 introduces the greenfield business model and isting infrastructure, staff, clients, or portfolios, and (ii ) the the landscape in SSA. It also describes the sample of green- central organizing bodies—often holding companies—that field MFIs and holding companies that participated in the create these MFIs through common ownership and manage- research for this publication. Detailed performance data ment. The holding company usually also plays a strong role in were obtained from 10 holding companies on 30 greenfield backstopping operations, providing standard policies and pro- MFIs in SSA. cedures, and co-branding subsidiaries in the network. Section 2 provides an analysis of the operational and fi- The greenfield model has come a long way in a short time nancial performance of greenfield MFIs with a focus on in SSA from seven greenfield MFIs in 2006 to 31 by 2012. their lifecycle, outreach, funding structure, and financial These are spread over 12 SSA countries, including post-con- performance. The life cycle of greenfield MFIs can be di- flict markets such as the Democratic Republic of Congo vided in three stages: foundation (preparation and first year (DRC), Cote d’Ivoire, and Liberia. While there is a range of of operation), institutional development (year two through microfinance providers in SSA, the proliferation of green- financial breakeven, which typically occurs in year three, field MFIs expands the commercial end of the spectrum with four, or five), and scale-up (from financial breakeven on- regulated, mostly deposit-taking institutions focused on mi- ward). The performance of greenfield MFIs largely reflects croenterprises and small businesses. At the end of 2012, the these three stages, each of which is characterized by mile- 31 greenfield MFIs in SSA had more than 700,000 loan ac- stones related to management, product development, infra- counts, an aggregate loan portfolio of $527 million, and close structure build out, outreach, funding structure, and sus- to 2 million deposit accounts with an aggregate balance of tainability. The average initial funding package required for $445 million. While many greenfield MFIs are still young, a greenfield MFI ranges from $6 million to $8 million over there are signs of solid institution building for the longer the first 3–4 years of operations and comprises a combina- term. At the end of 2012, they already employed more than tion of equity, technical assistance, and debt. Starting equi- 11,000 staff and had 700 branches. The greenfield MFIs are ty capital is about $3.5 million and is supplemented by tech- becoming noteworthy collectively and, in some cases, indi- nical assistance. For greenfield MFIs in the study, the vidually in their markets. technical assistance budgets ranged from $1.5 million to $9 The time is right to look more closely at how these MFIs million, but clustered around $4 million. Shareholders and and their holding companies build retail capacity, promote donors typically raise $3 million in grant funding on aver- market development, and ultimately advance access to fi- age, while the MFI pays the rest. 1 A comparison to Microfinance Information Exchange tures together with market share were key factors consid- (MIX) data for “young” African MFIs (those with 4–7 years ered. Particularly in the less developed financial markets, it of operations)1 shows that greenfield MFIs have achieved, on appears that greenfield MFIs play a pioneering role in ex- average, very robust performance. By the time greenfield panding the financial access of microenterprises, small busi- MFIs reach 60 months of operations, they have attained con- nesses, and low-income households. Their sustainable per- siderably larger size, greater reach, higher loan quality, and formance illustrates to the traditional formal banking sector better profitability than MFIs with no strong holding/net- that underserved businesses and households are bankable work affiliation. The average greenfield MFI tends to be and even profitable market segments. Additionally, green- much better capitalized and to have more formal structures field participation in credit bureaus, when available, helps to and deposit-taking infrastructure than the average young build the foundation for a strong credit culture and promotes MFI reporting to MIX. responsible finance for the market as a whole. The most im- Section 3 provides an overview of the structure and typol- portant effect on market building has come from greenfield ogy of holding companies that participated in the research, MFI investment in staff training and development. including how they are governed and funded and how they In the Conclusion, the authors contemplate the future implement the creation of greenfield MFIs. The typology role of greenfield MFIs in SSA. It is likely that the rate of cre- distinguishes between holding companies that are led by ation of greenfield entities, at least in SSA, will slow, as the consulting firms (Access Microfinance Holding, Advans, Mi- most “feasible” markets have now largely been entered. Yet, croCred, Procredit, Swiss Microfinance Holding) and those there remain about 25 countries in SSA without any green- that are helped by network support organizations (ASA, field MFI presence, and typically without the presence of BRAC, FINCA, and Opportunity International) and a com- any sustainable MFIs at all. And in almost every country, mercial bank (EcoBank). Holding companies and their peri-urban and rural populations still struggle to access fi- shareholders apply the greenfield business model to address nancial services. One challenge for greenfield MFIs and their some of the primary challenges in advancing access to fi- holding companies is, therefore, to develop a delivery model nance in SSA. The small size of nascent financial markets, that facilitates commercially viable and affordable access in high costs of doing business, uneven regulatory frameworks, smaller, more dispersed markets and rural areas. Alternative and inadequately skilled human resources in many SSA mar- delivery channels (including agent networks, mobile finan- kets benefit from an approach where practices can be stan- cial services, and related partnerships2) are an area of very dardized and costs can be shared. Through network struc- large investment for greenfield MFIs as they enter the scale- tures and common practices, holding companies are able to up phase. The greenfield MFI model is a complement to transfer knowledge and learning from one greenfield MFI to other strategies for increasing access to finance in SSA, such another. Their structured approach and heavy focus on hu- as reform of existing institutions without a holding structure man resources development have been a key success factor and bank downscaling. The next few years will be very tell- in their ability to create sustainable institutions in some of ing about the ability of greenfield MFIs to leverage their the most frontier markets in SSA. foundation and achieve scale, and for holding companies to Section 4 explores the contribution of the greenfield mod- replicate and sustain the success of their model in other mar- el in market development in three markets: the DRC, Ghana, kets, particularly in a context of diminishing funding for and Madagascar. While it is difficult to attribute changes in a technical assistance. Undoubtedly they will find themselves market or behavior of competing institutions to the interven- compelled to develop new methods, capacities, and practices tion of one or more greenfield MFIs, the authors used quan- to stay relevant and competitive in the microfinance space. titative and qualitative information (looking at market share At the same time, it is also likely that many of these green- and observed quality of services) to discern effects on the field MFIs will increasingly begin to compete with commer- overall level of financial inclusion and aspects of market cial banks for mass market customers and those in the small building. Development of human resources in the financial and medium enterprise space. The financial landscape in Af- sector and innovations in new products and/or product fea- rica is poised to become much more interesting.  IX index for “young” MFIs in Africa comprises 58 institutions between four and seven years old in December 2011. All but five of these institutions are 1. M deposit taking, with a deposit base ranging from $30,000 to $22.6 million ($2 million on average) and gross loan portfolio ranging from $2,000 to $24 million ($2.7 million on average). The greenfield MFI subjects of this paper were removed from the benchmark population. 2. F or more information see Flaming et al. (2013). 2 1 SECTION Introduction S ub-Saharan Africa (SSA) has the lowest level of access to banks. This cohort represents 90 percent of all greenfield finance of any region in the world, with an average MFIs created in Africa between 2000 and 2012. banked population of only 24 percent (Findex 2012). The holding company model. This section provides an The region’s banking systems are small in both absolute and overview of the structure and typology of holding compa- relative size (as measured by liquid liabilities and credit as per- nies, including how they are governed and funded and how centage of gross domestic product) (Beck, Maimbo, Faye, and they implement the creation of greenfield MFIs. The sec­ tion Triki 2011). The microfinance sector has been relatively slow is based on interviews with the management of the hold- to expand in SSA compared to other regions in the world. Ac- ing companies and review of annual reports and financial cording to the Microfinance Information eXchange (MIX) ­statements. landscape data, services are concentrated in larger urban cen- Contribution of greenfield MFIs to market develop- ters, and service delivery in rural areas is meager (MIX 2011). ment. This section explores the contribution of the green- Until a few years ago, the main providers of financial services field model in market development in the Democratic Re- to base-of-the-pyramid customers were credit unions, savings public of Congo (DRC), Ghana, and Madagascar. These and loans associations, and nonprofit credit programs. Now, countries were selected because they each have a critical new players are entering the region, including specialized mass of greenfield MFIs, most with a reasonably long track greenfield microfinance institutions (MFIs), downscaling Pan record. While it is difficult to attribute changes in a market or African commercial banks, and mobile network operators. behavior of competing institutions to the intervention of one This paper explores the greenfield business model, which or more greenfield MFIs, the authors used quantitative and focuses on expanding financial services through two main qualitative information (looking at market shares and per- elements: (1) creation of a group of “greenfield MFIs” de- ception of quality of services) to discern effects on the overall fined as institutions that are newly created without pre-ex- level of financial inclusion, the human resources skills base isting infrastructure, staff, clients, or portfolios, and (2) cen- in the financial sector, innovations in new products and/or tral organizing bodies—often holding companies—that product features, and the demonstration effect from the in- create these MFIs through common ownership and manage- troduction of and adherence to internationally recognized ment. The holding company usually also plays a strong role good practices, e.g., those related to transparency and client in backstopping operations, providing standard policies and protection. procedures, and co-branding the subsidiaries in the network. Given these commonalities, this model can also be consid- 1.1 Landscape ered a type of franchise where the sponsors inject a tested In SSA, the greenfield model made its debut in 2000 when approach and sufficient patient capital to move new institu- ProCredit Holding opened a bank in Mozambique (see Box tions past the difficult start-up phase and onto a growth tra- 1). For a few years, ProCredit was essentially alone in pursu- jectory in some of the most challenging markets. Three as- ing this strategy; it opened up in Ghana in 2002, in Angola in pects of the greenfield business model are addressed: (i ) 2004, and in the DRC in 2005. While other network opera- performance of subsidiary greenfield MFIs, (ii ) the holding tors and local institutions started nongovernmental organi- company model, and (iii ) contribution of greenfield MFIs to zations (NGOs) and cooperative microfinance entities much market development. earlier, the greenfield model of a centralized holding com- Performance of subsidiary greenfield MFIs. This sec- pany providing investment and expertise for the develop- tion analyzes the operational and financial performance of ment of commercial microfinance entities began in earnest greenfield MFIs with a focus on the lifecycle, outreach, fund- at the turn of the millennium. ing structure, and time required to break even. Detailed per- Between 2005 and 2006, Advans, Access, and MicroCred formance data were obtained from 11 holding companies on holding companies were formed with a structure similar to 30 greenfield MFIs in SSA, and interviews were conducted that of ProCredit and by the end of 2007 had collectively with more than a third of the chief executive officers of these launched five greenfield MFIs in SSA. Accion started its first 3 BOX 1 ProCredit Holding The first greenfield MFI was established by Internationale IPC created the investment company Internationale Mi- Projekt Consult (IPC) in Bosnia in 1996, and was licensed as a cro Investitionen AG (IMI) with shareholders, including IPC, deposit-taking microfinance bank. IPC had gained a lot of IPC staff, and development investors. The company became experience in institutional strengthening of microfinance pro- the main vehicle for expanding the successful Bosnian ex- viders by upscaling cooperatives and MFIs in Latin America periment, first to other Eastern European countries and then and downscaling banks in Uganda and Russia. Not satisfied to Latin America. In 2000 ProCredit expanded to SSA, with with being just the consulting company on these projects, the encouragement of its DFI shareholders. ProCredit set up IPC decided to invest in the institutions it was strengthening five banks in SSA, in Angola, DRC, Ghana, Mozambique, or creating. The main reasons for investing in greenfield MFIs and Sierra Leone. Since then it sold its banks in Angola and at the time were (i ) to have more control over capacity build- Sierra Leone and continues to have a network of 21 banks, ing and the growth trajectory of institutions, (ii ) encourage- of which three are in SSA. Over the past few years ProCredit ment from development finance institutions (DFIs) for IPC to has taken a global decision to change its target clientele “put some skin in the game,” and (iii ) a way for IPC to invest moving from microenterprises toward very small and small in the long-term value of the company. businesses. Source: Interviews and B. Fritz and K. Hujo (2005) greenfield MFI in the same period in partnership with three accounts, with an aggregate loan portfolio of $57.4 million, commercial banks in Nigeria. As a result of this initial experi- and held 220,377 deposit accounts with an aggregate balance ence, Ecobank and Accion entered into a partnership and of $50.7 million. Six years later, at the end of 2012, there were opened two greenfield MFIs in Ghana and Cameroon. From 31 greenfield MFIs3 in 12 SSA countries, with 769,199 loan that point on, the Access, Advans, and MicroCred networks accounts and an aggregate loan portfolio of $527 million, and each created more or less one new MFI per year. Toward the with 1,934,855 deposit accounts and an aggregate balance of end of the decade, ASA and BRAC from Bangladesh created new organizational structures that also allowed them to be- gin establishing greenfield MFIs in Africa. Over the six years FIGURE 1 from late 2006 to end of 2012, a total of 27 additional green- field MFIs were launched. SSA Countries in Which Greenfield MFIs were Meanwhile, FINCA and Opportunity International (OI) Created Between January 2000 and June 2012 have used the holding company structure to upgrade their ex- isting (largely NGO) affiliates to regulated deposit-taking in- stitutions, and then integrate them into a common investment company. Like the other networks, the holding company has been a vehicle for mobilizing investment capital, expanding and backstopping operations, and establishing an ownership Senegal model for an international network of financial institutions. Côte Ghana Nigeria Sierra Leone There is a range of strategies for extending the reach of d’Ivorie Liberia Cameroon Uganda microfinance, including the transformation of existing insti- Democratic Republic of Congo tutions, the creation of stand-alone greenfield MFIs without Tanzania a centralized management or holding structure, bank down- Angola scaling, and others. Zambia Madagascar When greenfield MFIs expanded in earnest at the end of Mozambique 2006, the first seven (Procredit Angola, ProCredit DRC, Pro- credit Mozambique, Procredit Ghana, FINCA DRC, Oppor- tunity Ghana, and MicroCred Madagascar) had 107,887 loan  xcluding the former Procredit companies in Sierra Leone and Angola, which were sold in 2007 and 2010, respectively. Technically, there have been 3. E several additional greenfield MFIs launched between June 2012 and the publication of this paper (ASA Kenya, Tanzania and Uganda, Oxus DRC, Advans Nigeria), but they were too recent to include. 4  rowth TABLE 1 G of Greenfield MFIs in SSA, 2006–2012 2006 2007 2008 2009 2010 2011 2012 Greenfield MFIs 7 12 18 22 27 30 31 No. of Staff 1,564 2,512 4,856 6,685 8,009 10,137 11,578 No. of Branches 37 56 261 392 514 625 701 No. of Loans Outstanding 107,887 141,231 332,349 449,973 570,017 743,640 769,199 Gross Loan Portfolio ($ million) 57.4 94.7 144.5 203.6 285.8 409.5 527.0 No. of Deposit Accounts 220,377 317,943 595,008 780,497 1,050,087 1,574,750 1,934,855 Total Deposit Balance ($ million) 50.7 106.7 177.9 211.6 291.3 371.8 445.5 $445 million. These 31 greenfield MFIs had 11,578 staff and ing did not provide performance data for these banks by 701 branches. These greenfield MFIs are becoming notewor- institutional age, but allowed the authors to use publically thy collectively (See Table 1), and in some cases individually, available data for the aggregate figures as well as the data the in their markets which is further discussed in Section 4 on authors had already collected for ProCredit DRC and Pro- the role of greenfield MFIs in market development. Credit Sierra Leone.4 IFC has direct investments in six of the holding companies and 17 greenfield MFIs in this cohort. 1.2 Sample Cohort Greenfield MFIs include a variety of legal forms. A major- According to the definition used in this paper, there were ity are licensed and regulated deposit-taking institutions, 33 greenfield MFIs created before June 2012 (see Table 2). ranging in legal structure from commercial banks to savings Institutions created through takeovers and mergers were not and loan companies to specialized deposit-taking MFIs. included, nor were many of the subsidiaries of OI and Some started as credit-only companies and relicensed as de- FINCA, which started as NGOs without strong central posit-taking institutions a couple of years after they were network bodies or holding company structures. created (FINCA DRC, MicroCred Madagascar); a few re- The detailed performance analysis of the greenfield MFIs main credit-only companies (the four BRAC entities). (Section 2) is based on information from 30 MFIs out of 33 The holding companies in this study represent a range of (see Table 2). These MFIs belong to the 10 holding compa- organizational structures that are explained in more detail in nies listed in Table 3. ProCredit Angola, Ghana, and Mozam- Section 3. Regardless of their structures, the holding compa- bique are not included (shaded in Table 2). ProCredit Hold- nies play an important role in identifying new markets and  reenfield TABLE 2 G MFIs Created in SSA Between January 2000 and June 2012 Greenfield MFI Country Start of Operations Current License (June 2012) 1 ProCredit Mozambique Mozambique 2000 Commercial Bank 2 ProCredit Ghana Ghana 2002 Savings and Loan Company 3 FINCA DRC DRC 2003 Deposit-taking MFI 4 ProCredit Angola Angola 2004 Commercial Bank 5 Opportunity Ghana Ghana 2004 Savings and Loan Company 6 ProCredit DRC DRC 2005 Commercial Bank 7 MicroCred Madagascar Madagascar 2006 Commercial Bank 8 Access Madagascar Madagascar 2007 Commercial Bank 9 Advans Cameroon Cameroun 2007 Deposit-taking MFI 10 Accion Nigeria Nigeria 2007 Microfinance Bank 11 MicroCred Senegal Senegal 2007 Deposit-taking MFI Continued  ata for ProCredit Sierra Leone are included only for the period that it was owned by ProCredit from 2007 until 2010, when it was sold to Ecobank. 4. D Procredit Angola was sold in 2007. 5  reenfield TABLE 2 G MFIs Created in SSA Between January 2000 and June 2012 cont’d Greenfield MFI Country Start of Operations Current License (June 2012) 12 ProCredit Sierra Leone Sierra Leone 2007 Commercial Bank 13 Access Tanzania Tanzania 2007 Commercial Bank 14 EB-Accion Ghana Ghana 2008 Savings and Loan Company 15 Advans Ghana Ghana 2008 Savings and Loan Company NGO (transformed into a Savings and 16 ASA Ghana Ghana 2008 Loan Company in April 2013) 17 Access Nigeria Nigeria 2008 Microfinance Bank 18 BRAC Tanzania Tanzania 2008 Credit-only Company 19 BRAC Uganda Uganda 2008 Credit-only Company 20 Advans DRC DRC 2009 Commercial Bank 21 Access Liberia Liberia 2009 Commercial Bank 22 BRAC Liberia Liberia 2009 Credit-only Company 23 ASA Nigeria (ASIEA) Nigeria 2009 NGO–MFI 24 BRAC Sierra Leone Sierra Leone 2009 Credit-only Company 25 Opportunity DRC DRC 2010 Deposit-taking MFI 26 MicroCred Ivory Coast Ivory Coast 2010 Deposit-taking MFI 27 EB-Accion Cameroon Cameroon 2010 Deposit-taking MFI 28 ASA Lagos (ASHA MFB) Nigeria 2010 Microfinance Bank 29 MicroCred Nigeria Nigeria 2010 Microfinance Bank 30 Fides Senegal Senegal 2011 Deposit-taking MFI 31 Advans Tanzania Tanzania 2011 Commercial Bank 32 Access Zambia Zambia 2011 Commercial Bank 33 Advans Ivory Coast Ivory Coast 2012 Deposit-taking MFI countries for expansion, funding and capital structure, and Now the time is right to look more closely at how these adhering to the vision and mission of the network. Each of MFIs and their holding companies build retail capacity, the holding companies has a cadre of specialists who provide promote market development, and ultimately advance technical assistance (TA) to the MFIs, help manage the hold- access to finance in SSA. A good number of greenfield ing company operations, and in many cases are founding in- MFIs now have a sufficient track record to enable an anal- vestors in the holding company. DFIs, socially responsible ysis of their performance and role in the market. A stock- investors, and specialized microfinance investment vehicles taking of the experience should help inform decisions (MIVs) are the primary investors in the holding companies that will shape the coming generation of investment in and greenfield MFIs. ­microfinance. 6  olding TABLE 3 H Companies Holding Company Sponsor Year Investors # MFIs # GF MFIs Total Assets of Created Globally in Africa Greenfield MFIs (June 2012) (June 2012) in SSA (US$, Dec 2012) Access Microfinance LFS 2006 CDC Group plc, European 7 5 157,966,796 Holding AG Investment Bank (EIB), International Finance Corporation (IFC), KfW Entwicklungsbank (KfW), the Netherlands Development Bank (FMO), LFS Financial Systems GmbH, MicroAssets GbR (MA), Omidyar-Tufts Microfinance Fund (OTMF) Advans SA SICAR Horus 2005 EIB, CDC, FMO, IFC, KfW, Horus 6 5 74,072,517 Development Finance, FISEA (Proparco) ASA International ASA 2006 Catalyst MF Investors (Owned by 8 3 37,060,395 Holding International Gray Ghost MF Fund, Sequoia, ABP, TIAA-CREF, CDC Group, responsAbility, private investors) BRAC International BRAC NGO 1972 BRAC International 6 4 60,006,805 Holdings, BV EcoBank International EcoBank 1985 Government Employees Pension 3 3 37,633,098 International Fund, Asset Mgt Corp of Nigeria, and Accion Social Securuty and National Insurance Trust, Stanbic Nominees Nigeria, IFC and IFC capitalization fund, Interlink Securities ltd. FINCA Microfinance FINCA 2011 FINCA International, IFC, KfW, 21 1 33,854,511 Holding Company LLC International FMO, responsAbility, Triple Jump MicroCred SA PlaNet 2005 PlaNet Finance, IFC, Société 6 4 24,893,865 Finance Générale, AXA Belgium, French Development Agency (AFD), EIB, Developing World Markets (DWM) Opportunity OI 2000 OI 15 2 52,027,591 Transformation Investment ProCredit Holding AG & IPC 1998 IPC GmbH, IPC Invest GmbH & 21 3 168,872,564 Co. KGaA Co., KFW, DOEN, IFC, BIO, FMO, TIAA-CREF, responsAbility, PROPARCO, Fundasal, Omidyar- Tufts Swiss Microfinance Fides 2007 Financial Systems Development 2 1 4,482,212 Holding, SA Services AG (Fides), Sobelnat, P.G.C. Suisse, Bank im Bistum Essen 7 8 2 SECTION Greenfield MFI Performance T he life cycle of greenfield MFIs, as it is observed in SSA, In the following sections, the performance of the cohort can be divided in three stages (see Box 2): (i) foundation of greenfield MFIs is evaluated based on institutional age to (preparation and first year of operation), (ii) institu- achieve a coherent comparison and aggregation of data, re- tional development (year two through financial breakeven, gardless of the calendar year in which they launched opera- which typically occurs in year three, four, or five), and (iii) tions. Data are presented as simple averages, unless other- scale-up (from financial breakeven onward). The perfor- wise indicated, to display the performance of a typical mance of greenfield MFIs largely reflects these three stages, greenfield MFI. In a few cases, when outliers distort the each of which is characterized by milestones related to man- simple average, a weighted average is used. agement, product development, infrastructure build out, out- A comparison to MIX data for “young” African MFIs reach, funding structure, and sustainability. (those with 4–7 years of operations)5 shows that greenfield BOX 2 Three Stages of Greenfield MFIs’ Life Cycle Foundation. The foundation stage includes the legal cre- growth. Eleven of the 30 institutions were in the institutional de- ation of the new entity, shareholder negotiations, the licensing velopment stage at the time of this analysis. Typically the MFIs process, and onsite operations preparation. Five of the 30 in- have only one product, group, or individual microenterprise stitutions included in this analysis are currently at this stage. In loans, but in some cases small and medium enterprises (SME) most jurisdictions a company needs to be created and partly lending is piloted. As operations grow, risk management systems capitalized before a preliminary approval can be sought from are increasingly institutionalized, including policies and proce- regulatory authorities. At that time initial management of the dures for decentralized management, internal audit, cash and li- greenfield MFI is also selected, usually consisting of staff sec- quidity management, and regulatory compliance, including anti- onded from the holding company. money laundering measures. The asset–liability committee at the The initial staff is responsible for tailoring policies and proce- board becomes more active as deposits increase and begin to dures to the local market, designing and adapting products, account for a greater portion of funds for intermediation. installing an information technology (IT) system, identifying and Scale-up stage. As greenfield MFIs pass breakeven and be- building or refurbishing physical space for branches, and man- come profitable, they move into a scale-up phase. Fourteen of aging the relationship with regulatory authorities. They also re- the 30 institutions in this study were in the scale-up phase at cruit one or two cohorts of loan officers (usually 20–30 in total) the time of the analysis. At this point, the focus tends to shift and train them for several months, often with network MFIs in toward product diversification and delivery channel develop- other countries. It usually takes 4–6 months to prepare for op- ment to attract new clients as well as deepen existing client erations from the point of receiving the preliminary approval relationships and gain market share. New products are devel- from the regulator. It can then take 2–4 more months until the oped to target secondary market segments, for example, agri- central bank inspects the greenfield MFI and grants the final cultural lending for rural clients. The expansion of small and operating license. Delays in the foundation stage can easily medium enterprise (SME) lending, in particular, can be a critical lead to substantial cost over runs, particularly if the holding driver of profitability by off-setting the high cost of smaller mi- company has placed staff on the ground too quickly. croloans as institutions expand their footprint into more rural Institutional development. During the institutional develop- areas. Greenfield MFIs have recently introduced automated ment stage, greenfield MFIs (in partnership with holding compa- teller machine (ATM) channels, and some institutions are now nies) focus on building staff capacity and installing risk manage- rolling out the first phase of various forms of agent networks ment systems that will create the core foundation for future enabled through point-of-service, cards, and mobile devices.  IX index for “young” MFIs in Africa, comprised 58 institutions 4–7 years old in December 2011. All but five of these institutions are deposit taking, with 5. M a deposit base ranging from $30,000 to $22.6 million ($2 million on average) and gross loan portfolio ranging from $2,000 to $24 million ($2.7 million on average). The greenfield MFI subjects of this paper were removed from the benchmark population. 9 TABLE 4 P  erformance of Greenfield MFIs at 12, 36, and ly-stage losses, and yet be sufficient to attract lenders who 60 Months a can provide debt funding. On average, greenfield MFIs in Month Month Month MIX Young SSA have started operations with equity capital of approxi- 12 36 60 Africa* mately $3.5 million. While the equity is typically partly No. Staff 131 318 524 69 eroded over the first 24–48 months as the MFIs make loss- No. Branches 9 22 31 10 es, Table 5 shows a moderately increasing equity level over No. Deposit-Taking 3 7 11 n/a time. This is explained by the fact that shareholders usually Branches inject additional equity to support the solvency and growth No. Nondeposit- 22 47 63 n/a of the MFIs, demonstrating their generally strong commit- Taking Branches ment to the endeavor. These capital injections have also No. Loans Outstanding 9,495 25,009 36,714 11,255 been a response to increasing minimum capital require- Gross Portfolio ($ million) 2.3 9.2 20.0 2.7 ments in many African countries during the past six years No. Deposit Accounts 7,123 37,460 81,682 18,127 and the realization among investor/sponsors that $3.5 mil- Deposit Volume ($ 0.8 8.7 23.1 2.0 lion has not generally been sufficient to absorb losses, com- million) ply with regulatory requirements (for banks), and raise PAR30 3.9 4.0 3.4 9.5 enough debt funding to support loans to the point of break- Op. Expenses / Avg 200 53 36 113 even. Initial capitalizations of around $4 million to $5 mil- Portf (%) lion are now more common. In comparison, the average Equity ($ million) 3.6 4.3 6.6 1.2 equity of the MIX Young Africa MFIs is $1.2 million after Net Income / Avg Assets -12.4 -0.1 3.1 -2.4 4–7 years. (%) In addition to equity, most institutions start with exter- Net Income / Avg Equity -44.6 -0.3 18.9 -3.4 nal TA grants of $3 million on average. Without this fund- (%) ing, the institution would have to pay for TA out of its own *n=58 African MFIs between 48 and 84 months funds (most greenfield MFIs in fact pay for some of the a. The figures in this table represent simple averages, except for net income/assets and net income/equity; flow data in ratios (e.g., operating expenses/average portfolio) are overall TA costs themselves, as will be explained below), based on annualized six-month data. incurring higher start-up losses that in turn would be ab- sorbed by the equity through increased retained losses. MFIs have achieved, on average, very robust performance. Taking into account both the initial equity capitalization By the time greenfield MFIs reach 60 months of operations, and the TA funding, the average initial funding package re- they have attained considerably larger size, greater reach, quired for a greenfield MFI ranges from $6 million to $8 higher loan quality, and better profitability than MFIs with million over the first 3–4 years of operations. But that’s just no strong holding/network affiliation. The average green- to get started. Naturally the MFI will soon also need addi- field MFI tends to be much better capitalized and to have tional funding in the form of debt and/or deposits to sup- more formal structures and deposit-taking infrastructure port a growing asset base. than the average young MFI reporting to MIX. (See Table 4.) The TA aspect plays a significant enough role in the fi- nancial and operational development of the greenfield 2.1 Balance Sheet Indicators MFIs to warrant a more detailed explanation (see, also, Box An early key decision for investors in greenfield MFIs is 3). The start-up TA program is designed to develop and how much to provide in start-up equity capital. The capital build capacities of the institution. The average TA program needs to comply with regulatory requirements, absorb ear- includes a six month to one year preoperational phase and Balance TABLE 5  Sheet Indicators of Greenfield MFIs Month 12 Month 18 Month 24 Month 30 Month 36 Month 42 Month 48 Month 54 Month 60 Assets ($ million) 6.0 8.5 11.4 13.6 16.5 20.9 25.8 28.0 34.9 No. in sample 29 28 27 23 22 20 17 14 13 Equity ($ million) 3.6 3.5 3.6 4.0 4.3 4.5 5.3 5.3 6.6 No. in sample 27 27 26 22 21 20 17 14 13 Equity / Assets (%) 59 47 39 36 33 29 24 23 24 No. in sample 29 28 27 23 22 20 17 14 13 Deposits / Loans (%) 25 36 40 43 42 43 43 53 60 No. in sample 24 23 22 18 17 16 15 12 11 10 a 36–48 month operating phase intended to support the in- costs. The institution contributes the balance, usually 20– stitution through the foundation and institutional develop- 40 percent of the total. For greenfield MFIs in this cohort, ment ­ stages. The TA budget typically funds 3–4 long-term the TA budgets range from $1.5 million to $9 million, but senior managers, installation and tailoring of IT application cluster around $4 million. In those cases, shareholders and and management information systems, several short-term donors typically raise $3 million in grant funding on aver- specialists in internal audit, risk management and product age, leaving the MFI to pay the rest. development, and technical backstopping from the holding As expected, greenfield MFIs typically remain modestly company/technical service provider. Once the TA budget is leveraged over the first 60 months of operations. They start established, shareholders and donors typically try to pro- out with an equity-to-asset ratio at 100 percent, though some vide enough grant funding to defray a majority of the TA of it is eroded by preparation activities and by initial losses BOX 3 Impact of Technical Assistance on Financial Performance Why do greenfield MFI projects generally include sub- field MFIs receive on average $3 million in external TA grants stantial amounts of TA grant funding? Fundamentally, it is for their start-up period. In addition, they typically pay about related to internal constraints of investors, such as risk-re- $1 million out of their own pocket, for a total TA budget of $4 turn preferences and time horizons, and the belief among million. Table B3-A attempts to illustrate what would happen donors and investors (some investors provide both equity if the full TA cost were borne by the MFI. and TA funding) that there are potentially broad benefits to This simulated example shows that typical greenfield MFIs well-run and (eventually) large MFIs that can offer a mean- would experience higher retained losses if paying fully for the ingful range of financial services to microenterprises, small TA. There is also a lot more volatility in the return on average businesses, and low-income populations in SSA (elaborated equity (ROAE), fueled by higher initial losses and diminished in Section 4 on market development effects). equity. The time to reach the monthly breakeven point, how- On the first point (internal investment constraints), it is of ever, remains the same at month 42. But since the retained course possible that greenfield MFIs may turn out to be losses are higher and will take longer to recover, the expected good investments for the initial investors. However, it is un- return to investors is lower. Without TA grants, the expected likely that this will happen over any reasonable time hori- internal rate of return (IRR) at five years is approximately 1 zon, which most investors would consider to be 5–8 years. percent; with TA grants, it is approximately 14 percent. An IRR While development-oriented investors may accept lower of 1 percent is too low for DFI investors to justify an invest- expected returns for higher expected impact, they also face ment, even if they consider the investments to have an impor- limits to how far this can be stretched. tant development effect on the local market. In fact, 14 per- What does this burden-sharing look like, and how does cent is below what many DFIs and social investors would it affect the finances of the MFIs? As noted earlier, green- consider acceptable in a region like Africa. TABLE B3-A Example Calculation Month 12 Month 18 Month 24 Month 30 Month 36 Month 42 Month 48 Month 54 Month 60 Net Income for the period ($) (410,387) (359,666) (178,023) (33,527) (26,280) 174,318 419,463 553,862 395,553 Add'l Cost to MFI if no external TA funding ($) (500,000) (500,000) (500,000) (500,000) (500,000) Net Income if no external TA funding ($) (910,387) (859,666) (678,023) (533,527) (526,280) 174,318 419,463 553,862 395,553 Equity ($) 3,553,198 3,510,502 3,558,164 3,839,706 4,324,016 4,480,075 5,267,880 5,284,887 6,558,059 Equity if no external TA funding ($) 2,553,198 2,010,502 1,558,164 1,339,706 1,324,016 1,480,075 2,267,880 2,284,887 3,558,059 Annualized ROAE (%) -24.4 -23.7 -15.1 -5.8 -1.5 3.6 12.4 19.9 16.1 Annualized ROAE if no external TA funding (%) -71.5 -78.9 -74.8 -72.3 -73.5 -25.0 33.1 51.7 32.6 The table builds on the following methods and assumptions: (i) the actual average net income and equity positions for the greenfield cohort were used as a starting point; (ii) the $3 million received in external TA grants is spread evenly across the first 36 months of operations; (iii) the remaining $1 million funded by the MFI is already reflected in the average net income and equity figures. 11 TABLE 6 Average Balances of Greenfield MFIs Month 12 Month 18 Month 24 Month 30 Month 36 Month 42 Month 48 Month 54 Month 60 Avg. Loan Balance ($) 703 784 795 743 841 896 929 1,003 1,147 No. in sample 29 28 27 23 22 20 17 14 13 Avg. Deposit Balance ($) 152 214 238 202 228 251 245 251 207 No. in sample 23 22 21 17 16 15 14 11 9 Note: The average loan balance ($1,147) and deposit balance ($207) at month 60 reflect 101 percent and 22 percent, respectively, of gross domestic product per capita in 2011 of countries represented. unless shareholders inject additional equity. Over time, as their shareholders (mainly DFIs). profitability stabilizes, the equity-to-asset ratio is deter- Greenfield MFIs mobilize local resources in various forms mined more by loan growth and deposit mobilization. In the but, given their lack of a local track record, they initially cohort there is more variation among the greenfield MFIs depend significantly on debt funding from DFIs and created as nonprofit organizations or credit-only companies specialized MIVs managed by entities such as responsAbility, (sometimes the ratio becomes very low or even negative), Symbiotics, Blue Orchard, Triple Jump, and MicroVest. presumably because these entities are not as bound by for- Local banks have so far been reticent to fund greenfield mal regulations regarding minimum capital and possibly also MFIs. A transition to local deposit and local debt funding is because of capital constraints at the sponsor level. taking place with support from partial credit guarantees, The deposit-to-loan ratio, which is an important indica- etc., but this will require time: even after five years of tion of the ability of greenfield MFIs to gain the trust of lo- operations many greenfield MFIs battle to raise sufficient cal populations and raise stable resources for on-lending, funding locally to support their (typically) rapid loan growth. shows a continuous improvement over time. Nevertheless, the data in this paper clearly show that it is easier for green- 2.2 Growth and Operational Performance field MFIs to sign up depositors than it is to raise substan- Greenfield MFIs have generally achieved impressive tial volumes of deposits. The low-income nature of the de- growth, with the average MFI having 36,714 loans, $20 mil- positor base means that amounts are small and that lion loan portfolio, 81,682 deposit accounts, and $23.1 mil- greenfield MFIs typically have to rely on significant lion in deposit volume on the books at 60 months (see Table amounts of borrowings during their first 60 months of op- 7). Nevertheless, the ranges shown in figures 2, 3, 4, and 5 erations. See Table 6. The ability to mobilize deposits also hint at the diversity in lending and deposit mobilization depends on a couple of other factors, such as institutional strategies among greenfield MFIs. license (it tends to be easier for banks than nonbanks to Some greenfield MFIs focus on institution building and mobilize deposits), and local market conditions (e.g., the prioritize the development of a full menu of credit, savings, reputation of the microfinance industry or even the bank- and fee products early on, while others focus on expanding ing sector). In some cases, particularly over time, greenfield their footprint and prioritize the roll-out of a single loan MFIs can come to be perceived as safer than local banks product. For example, credit-only institutions focused pri- because of high standards of service and/or the nature of marily on group lending tend to reach a large number of TABLE 7 Growth Indicators of Greenfield MFIs Month 12 Month 18 Month 24 Month 30 Month 36 Month 42 Month 48 Month 54 Month 60 No. Loans Outstanding 9,495 12,504 18,622 23,045 25,009 28,346 32,554 35,569 36,714 No. in sample 29 28 26 23 22 20 16 14 13 Gross Portfolio ($ million) 2.3 4.0 6.0 7.2 9.2 11.6 15.6 17.4 20.0 No. in sample 29 28 27 23 22 20 17 14 13 No. Deposit Accounts 7,123 13,738 21,136 31,551 37,460 48,900 61,743 71,174 81,682 No. in sample 23 22 21 17 16 15 14 11 9 Deposit Volume ($ million) 0.8 2.6 4.4 6.1 8.7 12.2 14.2 17.5 23.1 No. in sample 24 23 22 18 17 16 15 12 11 PAR30 3.9% 4.3% 4.5% 4.4% 4.0% 3.7% 2.9% 3.7% 3.4% No. in sample 23 28 27 23 22 20 17 14 13 12 outstanding loans relatively quickly (though usually with a maintaining robust operational control. At 12 months of op- relatively modest overall loan portfolio volume). BRAC, in eration, greenfield MFIs have on average 131 staff and nine particular, has been very quick to scale up its lending pro- branches; at 60 months of operation, they have on average grams and has managed to reach more than 100,000 out- 524 staff and 31 branches. It is important to note, however, standing loans in Tanzania and Uganda within 30 months that the rate of branch expansion varies greatly between of operation. credit-only institutions and regulated deposit-taking insti- The growth numbers in Table 8 are a reflection of the tutions. It requires much more planning and investment, ability of greenfield MFIs to build out distribution net- and sometimes regulatory approval, to set up deposit-tak- works (mainly branches and outlets) and train staff while ing branches. Regulated deposit-taking institutions in the FIGURE 2 FIGURE 4 Evolution of Number of Loan Accounts Evolution of Gross Loan Portfolio ($US) 120,000 35,000,000 Gross Loan Portfolio by MFI ($US) 100,000 Number of loan accounts by MFI 30,000,000 80,000 25,000,000 60,000 20,000,000 15,000,000 40,000 10,000,000 20,000 5,000,000 0 0 6 12 18 24 30 36 42 48 54 60 6 12 18 24 30 36 42 48 54 60 Time after launch (months) Time after launch (months) Largest GLP Smallest GLP Median Largest number of loans Smallest number of loans Median FIGURE 3 FIGURE 5 Evolution of Deposit Accounts Evolution of Deposit Volume ($US) 120,000,000 140,000 Number of deposit accounts by MFI 100,000,000 120,000 Deposit volume by MFI ($US) 100,000 80,000,000 80,000 60,000,000 60,000 40,000,000 40,000 20,000 20,000,000 0 0 6 12 18 24 30 36 42 48 54 60 6 12 18 24 30 36 42 48 54 60 Time after launch (months) Time after launch (months) Largest number of accounts Smallest number of accounts Median Largest deposit volume Smallest deposit volume Median 13 TABLE 8 Staffing and Branches of Greenfield MFIs Month 12 Month 18 Month 24 Month 30 Month 36 Month 42 Month 48 Month 54 Month 60 Number of Staff 131 196 232 303 318 351 420 445 524 No. in sample 28 26 26 22 21 19 17 14 12 No. Branches 9 14 15 21 22 24 26 29 31 No. in sample 29 26 27 22 21 19 17 14 13 No. Deposit-Taking Branches 3 4 5 6 7 7 9 10 11 No. in sample 19 17 17 13 13 11 11 9 8 No. Nondeposit-Taking Branches 22 33 34 43 47 48 57 63 63 No. in sample 10 9 10 9 8 8 6 5 5 Loans / Staff member 65 59 73 73 68 72 72 71 71 No. in sample 28 26 26 22 21 19 17 14 12 cohort opened on average 11 branches in the first five years, quite a lot of variation in the greenfield MFI numbers, spe- whereas credit-led models opened 75. cifically between entities based on group lending (which Staff development is critical for sustained growth. Dur- tend to be credit only) versus those based on individual ing the first 3–4 years, successful loan officers are promoted lending (which tend to be deposit taking). Group lenders to supervisors, branch managers, and regional managers, tend to have significantly higher loans-to-staff ratios. slowly replacing international staff (typically there will be only one, perhaps two, international staff left at 48 months). 2.3 Financial Performance Several of the networks have created group-wide staff de- velopment programs that enable national staff to rotate to Financial performance is an integral aspect of the sister institutions to be trained or to train others. Staff ex- greenfield MFI model since many of its investors care changes are also used to support specific initiatives, such as almost as much about financial returns as they do about SME lending or the launch of debit cards. development impact. For these investors, it is important to Most greenfield MFIs recruit and train young adults see a steady progression toward financial sustainability who have little work experience and are new to the banking through rising revenues, falling cost ratios, and improving sector. Many new staff are recruited with basic high school margins and returns. math skills and are trained in cash-flow-based credit Indeed, Table 9 shows that the greenfield MFIs in the analysis and customer service, reflecting the qualities that cohort have been able to sustain fairly rapid revenue growth characterize the credit culture of most greenfield over their first 60 months, increasing on average by $500,000 institutions. Marketing largely focuses on bringing banking every six months and reaching $5 million by the five-year services to clients, rather than having clients come to the anniversary. At the same time, they have managed to push bank. As such, employees are selected for their ability to operating expense ratios lower. However, this has not meant relate to and communicate effectively with clients in that the trajectory to financial sustainability and profitability markets and at the place of their businesses. has been entirely smooth. The averages in the table give the Staff productivity levels have improved steadily among impression of a stable progression toward sustainability but, greenfield MFIs in SSA (as measured by the loans-to-staff in fact, greenfield MFIs typically experience significant ratio), but the cohort has nevertheless struggled to reach swings from profits to losses and back to profits during this the same productivity numbers as in other parts of the period.6 Many greenfield MFIs register substantial losses world. This is probably due to many greenfield MFIs hav- over the first 24 months before achieving initial breakeven ing a significant number of noncredit personnel involved in around 24–36 months but then, as they begin to assume the banking operations, their relatively stronger focus on SME full cost of any additional management service contracts, lending, and the continued rapid recruitment of new loan fall back into losses for the next 6–12 months. Only around officers. However, it should again be noted that there is months 42–48 do they emerge fully self-sustainable.  he authors have not attempted to remove the TA support from the figures presented in this paper because the amount of the support is very difficult to 6. T precisely quantify and attribute among different accounting periods. 14 Financial TABLE 9  Ratios of Greenfield MFIs Month 12 Month 18 Month 24 Month 30 Month 36 Month 42 Month 48 Month 54 Month 60 Total Revenue ($ million) 0.62 0.98 1.59 1.98 2.46 2.75 4.03 4.27 5.02 No. in sample 28 26 25 23 21 19 17 14 13 Portfolio Yield (%) 59 55 56 56 54 54 55 54 52 No. in sample 28 25 23 20 21 19 16 14 13 Op. Expenses / Avg Portf (%) 200 108 82 57 53 45 38 37 36 No. in sample 28 26 24 21 21 19 16 14 13 Net Income ($ million) (0.39) (0.35) (0.17) 0.01 (0.03) 0.17 0.42 0.55 0.40 No. in sample 28 27 26 23 22 20 17 14 13 Net Income / Revenue (%) -120 -69 -26 -13 -11 -5 10 12 8 No. in sample 28 26 25 23 21 19 17 14 13 Net Income / Avg Assets (%) -12.4 -8.8 -4.1 0.4 -0.1 1.8 3.3 3.8 3.1 No. in sample 28 27 26 23 22 20 17 14 13 Net Income / Avg Equity (%) -44.6 -24.2 -13.7 -0.3 -0.4 -3.9 20.0 26.0 18.9 No. in sample 28 27 26 23 22 20 17 14 13 Greenfield MFIs charge interest rates that typically generate portfolio yields around 55 percent for the first five FIGURE 6 years. The average operating expense ratio of greenfield MFIs shows steady improvement over time, falling to 36 Evolution In Portfolio Yield of Greenfield MFIs percent at 60 months of operation. However, both portfolio yields and operating expense ratios are high compared to 120 mature MFIs in other regions of the world, which ranged from 11 percent to 16 percent in 2011 (Rosenberg, Gaul, 100 Ford, and Tomilova 2013). To align performance with MFIs % Portfolio yield in other regions, the greenfield MFIs in SSA will need to 80 further reduce the operating expense ratio by 10–20 60 percentage points, something that may not be easy given the high costs of doing business in the region. See Figure 6. 40 Establishing new MFIs is a difficult endeavor, with many potential challenges and pitfalls. Some of the most common 20 mistakes and problems that greenfield MFIs have 0 experienced in SSA are highlighted in Box 4. 6 12 18 24 30 36 42 48 54 60 Time after launch (months) Highest portfolio yield Lowest portfolio yield Average 15 BOX 4 Common Challenges and Pitfalls in Establishing Greenfield MFIs The startup stage takes longer and costs more than Overly aggressive expansion. Greenfield MFIs sometime expected. The preoperational stage is a complex undertak- undertake aggressive expansion, driven by the holding company, ing that, among other things, includes interacting with reg- board, or management, in an effort to quickly reach a large loan ulatory authorities, identifying and renovating office space portfolio and financial breakeven. Such as the case for stand- for branch operations, recruiting and training staff, tailoring alone MFIs, this generally leads to high levels of nonperforming policies and procedures, and configuring the IT platform. If loans. Sometimes this mistake simply involves making too many cost overruns occur due to regulatory delays or poor plan- loans too quickly, but sometimes it also involves poorly managed ning or execution, less funding is available for the opera- or premature transition into SME lending before having built tional phase. basic internal expertise. In some cases, it also involves Key management staff leaves prematurely. The CEO establishing too many branches too quickly, which can also or other key staff sometime depart prematurely, in rare weigh on profitability through higher asset depreciation. instances before the duration of their initial contract, but Growth exceeds funding. Significant discrepancies are generally before the institution reaches the break-even often discovered between projected and actual demand, point or scale up phase, either due to difficult living particularly regarding deposit mobilization. Sometimes deposit conditions, family reasons, or disagreements with the board growth takes off immediately, but more often it lingers at low or the technical service provider. Sometimes the problem is levels for 15–20 months before taking on a more expected exacerbated by attrition of critical frontline staff, such as trajectory. Rarely does it follow a smooth curve in line with loan loan officers. Between 2006 and 2010 six CEOs of 12 portfolio growth. Clearly, a situation of slower than anticipated greenfield MFIs in SSA departed prematurely. Given that deposit mobilization can create a major bottleneck with regard management is one of the key success factors in greenfield to funding, so it is important that greenfield MFIs start out with MFIs, disruptions in the senior management team can one or two potential lenders closely associated with the venture create serious difficulties and significantly impact (some DFI shareholders, such as IFC and FMO, can also fill the performance. role of lender to greenfield MFIs). Source: CEO Interviews and Jansson (2010). 16 3 SECTION The Holding Company A ccording to the definition used in this paper, green- holding companies. In the consulting-firm-led model, the field MFIs belong to a larger network or holding sponsor (i.e., the consulting firm) typically holds a 3–20 per- company, which through common ownership and cent minority ownership of the holding company. However, management, plays a strong role in backstopping operations, in the NSO-led model, the sponsors typically have a much providing standard policies and procedures, providing staff larger stake in the holding company, often above 50 percent, development and training, and co-branding the subsidiaries as their shareholding in the holding companies came from in the network. This section explains the holding company contributing shares they held in a relatively large number of structure: who its investors are and how it is governed, its existing MFIs. BRAC and ASA are the only two sponsor orga- mission and key strategic choices in building out a network, nizations that have not invested in the holding companies, as and how management capacity is fostered and knowledge Bangladeshi law prohibits nonprofit organizations such as transferred. BRAC and ASA from owning shares in foreign companies, whether holding companies or operating companies. 3.1 Structure Sponsor management of the holding company. In sev- Each of the holding companies is very much an extension eral cases, the holding companies have been launched as in- of the historical operations and strategies of the group found- vestment companies with minimal if any administrative ers (the “sponsors”), reinforced by the investors who joined staff, managed by the sponsor organizations under a contract as initial shareholders. Three types of sponsor organizations with the holding company and its shareholders. This is the are present in the holding companies covered in this study: case for networks led by consulting firms (Advans, Access, specialized microfinance consulting firms, microfinance net- etc.), as well as some of the NSO-led ones (FINCA, Opportu- work support organizations (NSO), and a regional bank (see nity). The BRAC and ASAI holding companies have more in- Box 5).7 Despite their distinct origins, the holding companies house administrative staff, presumably because they must share several structural similarities, though there are some function independently from the sponsor organizations. interesting variations. Management and technical support to greenfield MFIs. Sponsor investment in the holding company. The spon- A primary role of the holding companies is to secure, directly sor organizations are normally founding investors in the or through an associated technical service provider, consis- BOX 5 Typology of Holding Companies Creating Greenfield MFIs Consulting firm led. Five of the holdings were founded NSO led. Four of the holding companies that create by specialized consulting firms for the purpose of investing greenfield MFIs were established to consolidate the affiliates in and building a global network of subsidiaries. These ini- of international microfinance networks and expand with new tial sponsors, all based in Europe, wanted to be more than greenfield MFIs. FINCA, BRAC, ASAI, and OI all belong to mere service providers; they wanted to be investors in this category. branded microfinance networks. The sponsor consulting Local bank led. Ecobank is a Togo-based bank (i.e., the firms and their related holding companies are LFS and Ac- holding company is based in Togo) with operations in 33 cessHolding, Horus and Advans, FIDES and Swiss Microfi- countries in SSA. It participated as a shareholder in Accion nance Holdings (SMH), and IPC and ProCredit. MicroCred Microfinance Bank in Nigeria and expanded its mass market was started by PlaNet Finance, a nonprofit organization operations by creating specialized MFIs in Ghana and with a consulting arm and relatively diverse microfinance Cameroun in collaboration with Accion International and interests. Accion Investments in Microfinance (AIM).  ome other large commercial banks that also provide microfinance services, such as Equity bank, have expanded to new countries. However, these banks, 7. S or their subsidiaries, do not focus exclusively on the MSE market and therefore have not been included in the study. 17 tent and high-quality support to the greenfield MFIs during 3.2 Investors their different lifecycle stages. In the case of the consulting- DFIs have played a key role in creating and supporting firm-led networks, this service is provided by the sponsor most of the networks that launch greenfield MFIs today. For organization, i.e., the consulting firm behind the network. most of the consulting-firm-led networks, the holding com- Several of the consulting-firm-led networks are considering panies began as a partnership between the sponsor and a (or may have already considered) migrating to an organiza- core group of DFIs: EIB, IFC, KfW, FMO, and AFD. The tional arrangement where the technical capacity of the spon- group of investors has somewhat expanded over time, but it sor organization is merged with the holding company, to is still dominated by DFIs (see Table 3). DFIs also played a achieve even closer alignment of interests and a simpler or- role in the transformation of the FINCA network into a hold- ganizational set-up that can more easily attract private inves- ing company model, contributing significant amounts of eq- tors in the future. uity and debt funding to enable the transformation and con- Ownership of the subsidiaries. The holding companies tinued expansion of their network MFIs. (See Box 6.) in this study generally aim to hold at least 50 percent owner- DFIs have actively supported the creation of the holding ship in their greenfield MFIs. There are a few African green- companies for several reasons. The holding company model field MFIs where this is not the case (e.g., Accion Nigeria and has provided DFIs with a single vehicle for making larger in- Fides Senegal) but these cases are rare, and becoming rarer vestments in microfinance and leveraging their participation still. The key people behind the holding companies are in- with other investors. The holding companies are also seen as creasingly focused on making sure that operational responsi- providing a relatively feasible exit route when DFIs believe bilities and financial incentives are appropriately aligned. their role has been completed, as shares in a geographically Alignment would be weaker if the holding company, which is diversified holding company are thought to be easier to sell responsible for providing technical and management servic- than multiple small investments in “difficult” countries. es to the greenfield MFIs (whether directly or through a Equally important, the leading DFIs wanted to create com- linked service provider), does not have a shareholder stake in mercial incentives to ensure the full engagement of the spon- the greenfield MFI that is appreciably larger than any of the sors. The holding company arrangement has engaged the other investors, preferably a majority stake. However, if the consulting firm and NSO sponsors as shareholders in the holding company has limited financial resources, it may face holding company, where they stand to gain or lose along with a trade-off between the degree of ownership control and the the other investors. number of entities it can launch. It may therefore accept to The second largest group of funders in the holding com- hold less than 50 percent, at least for an initial period, if the panies and the greenfield MFIs is socially responsible MIVs. minority shareholders are considered very like-minded. (See Organizations such as the Omidyar–Tufts Microfinance Figure 7.) Fund, Doen Foundation, Developing World Markets, re- sponsAbility, Triple Jump, Incofin, and Gray Ghost have all invested in the holding companies and/or in the greenfield FIGURE 7 MFIs. They tend to prefer the holding companies, for diver- sification and liquidity, and typically see their role more as Holding Company Structure providing expansion capital than venture capital. They therefore do not typically participate with equity in the Investors Shareholding founding stage of individual greenfield MFIs. However, they Investment sometimes enter during the institutional development or the Sponsor/TA Provider Holding Company scale-up stage. Several MIVs also provide debt funding to greenfield MFIs. Private commercial investors have shown interest in the Management holding companies as well. Large companies such as Axa, Services Lead TIAA-CREF, Sequoia, ABP, and Bank im Bustum Essen have Shareholder made significant investments. Small, individual investors are also present. MFI MFI MFI MFI It is possible to discern some different approaches to fund- Minority ing strategy and investor selection by the sponsors of the dif- Investment & TA Funding ferent networks. The consultant-led networks generally have Note: This illustrates primarily the consulting-firm led model. In the case of oth- ers, the sponsor/TA provider would be together in the same box as the holding DFI-dominated ownership, though Fides took a very differ- company. ent approach and sought out mainly individual investors with higher risk appetite and longer investment horizon than DFIs 18 NSO-led companies have taken a mixed approach. When BOX 6 FINCA created FINCA Microfinance Holding (FMH) in 2010 it focused heavily on identifying like-minded investors DFI Funding for Holding Companies willing to weigh social and financial performance equally. As and Greenfield MFIs such, FINCA International (NGO sponsor) maintained a ma- Greenfield MFIs and their related holding companies have jority stake in FMH, and a combination of DFIs and socially become an important vehicle for MFI investors. Many DFIs responsible investors (IFC, KfW, FMO, responsAbility, and support the holding companies at both the holding and the Triple Jump) invested in the remaining shares. OI formal- greenfield MFI levels. As of the end of 2011, 7 percent of ized its network in 1998 when all affiliated partner organiza- global DFI funding for microfinance ($676 million) went to tions signed a membership agreement with the OI network. greenfield holdings, up from 4 percent in 2007. OI programs are financed through direct solicitation of funds Most of the DFI investments at the holding level (87 per- from individuals, corporations, foundations, and religious cent) went to the consulting firm-led model, with the largest organizations. OI created Opportunity Transformation In- share going to ProCredit, while BRAC and FINCA essentially vestments (OTI), a wholly owned subsidiary, to invest in and received the other 13 percent in 2011. KfW, OPIC, and IFC provided three-quarters of all the funding, with most of the hold ownership positions in OI affiliates. remaining amounts coming from AFD, Proparco, FMO, EIB, ASAI and BRAC have chosen to partner mainly with MIVs BIO, and CDC. and institutional investors. ASAI is 100 percent owned by Catalyst Microfinance Investors (CMI), a Mauritius-based FIGURE B6-A DFI Funding (US$ million, Dec. 2011) company whose shareholders include more socially respon- sible private investors, pension funds, and MIVs than DFI Grant 11.3 11.8 funding. Much of the funding for BRAC subsidiaries comes from the BRAC Africa Loan Fund, a Cayman Island vehicle 17.9 Debt 119.1 established with funding from OPIC, MIVs, and socially re- Green eld MFIs sponsible investors specifically interested in supporting 21.5 Guarantee 192.6 Holdings BRAC’s expansion. 43.6 The case of Ecobank is naturally different, as the group Equity 353.3 was not created for the purpose of extending microfinance 0 100 200 300 400 services. Consequently, the investors of the holding company USD million do not represent the same focused interest in microfinance as can be found behind the other networks. Rather, they are fairly mainstream institutional investors with rather broad Greenfield MFIs represented more than a quarter of the agendas, raising the question of how microfinance will be ac- total number of all DFI direct investments in financial institu- commodated and supported in the larger organization. tions in SSA (25 of 91) at the end of 2011. Close to half of the All of the holding companies require at least a majority funding directed at greenfield MFIs was invested as equity in share in the network MFIs to exercise control, which they 25 entities. In terms of debt, nine received a loan and eight a consider to be a critical success factor. The consulting-firm- guarantee from a DFI. Eighty percent of the DFI funding went led holding companies own majority shares in the MFIs in to MFIs of the consulting-firm-led holdings and the other 20 partnership with minority investors, typically the same DFIs percent went to affiliates of FINCA, OI, and Accion. At the end of 2011, IFC, AFD/Proparco, and KfW accounted for 64 per- that have invested in the holding company. In the FINCA cent of all direct DFI funding to the individual greenfield MFIs. network, the holding company owns 100 percent of the Other public funders in addition to those invested in the hold- MFIs. OTI is the majority owner of the OI subsidiaries, and ings mentioned above include the African Development Bank the remaining shares are divided among the individual OI and DCA USAID (for guarantees). companies in Canada, Australia, and the United Kingdom. The BRAC and ASAI holding companies also strive to own Source: 2012 CGAP Cross-Border Funder Survey 100 percent of subsidiaries but have made exceptions where legally required. Ecobank has opted to establish subsidiary finance companies to clearly separate microfinance opera- when it created Swiss Microfinance Holding in 2007. DFIs tions from Ecobank’s core retail business. These subsidiaries may eventually be invited to support SMH’s next (expansion) are typically owned and controlled by the local Ecobank sub- phase. It is also important to note that in most consulting- sidiary together with the Ecobank International holding firm-led networks there is substantial financial participation company, but Ecobank has also included a few strategic part- by the individuals behind the sponsor companies. ners as minority investors, notably AIM and IFC. (See Box 7.) 19 BOX 7 Accion Investments in Microfinance Ecobank’s microfinance greenfielding ventures were Ghana and Cameroon.b Ecobank also participated as a undertaken in collaboration with AIM and Accion International. minority shareholder in a greenfield MFI in Nigeria launched The idea was to use Ecobank’s footprint in the region and—in by Accion. the creation of greenfield MFIs—and combine it with Accion The Ecobank–Accion model involves an important International’s technical expertise and AIM’s investment challenge in that the operational responsibility (which lies capital. Under this model, Accion International would provide initial capacity-building services through three-year TA with Accion International) and the financial responsibility programs, while AIM supported the ventures financially (which lies primarily with Ecobank) are not completely through significant minority equity stakes.a This approach aligned. It is to avoid this challenge that many holding suited all parties: it gave Ecobank a chance to roll out companies are closely linked with the technical service microfinance operations under experienced supervision; it provider and insist on taking majority stakes in the greenfield allowed AIM to invest with a 5–8 time horizon (which coincided MFIs. Over time, Ecobank and Accion will need to address with its fund life), and it enabled Accion International to this challenge, particularly if they intend to create additional expand its activities in Africa through a high-visibility greenfield MFIs. Lately Ecobank and Accion have pursued partnership with multiple replication opportunities. The new microfinance projects in SSA independent of one collaboration has produced two greenfield MFIs so far: in another. a. In 2012 the shareholders of AIM sold their shares to Bamboo Finance (a global private equity group specializing in microfinance and social entrepreneurship) and, in the case of AIM’s Africa portfolio companies, to Accion’s Gateway fund. b. While Ecobank has a majority stake in the projects in Ghana and Cameroon, the shareholding in Nigeria is more fragmented (it has a total of six shareholders). However, Ecobank is the largest minority shareholder with 24 percent, and Ecobank is an important investor with 18.5 percent. Governance. Governance for greenfield MFIs is meant to During the foundation stage, the shareholders agree on support long-term growth and development. Viewed as one articles of incorporation and the shareholders agreement, of the primary benefits of this model, high-quality and con- which detail the mission and governing principles of the in- sistent governance practices provide a strong foundation and stitution, board structure, and voting requirements. Majority enabling operating environment from inception and and super majority votes are generally stipulated for strate- throughout the institutions’ lifecycle. Particularly critical to gic decisions, including the issuance of new shares, entry of governance is the ownership structure and board composi- new investors, changes to primary operating policy guide- tion, which influence key strategic decisions related to the lines of the institution, and the procurement of technical and shareholders agreement, articles of association, and key op- management services. The large number of decisions requir- erating ­principles. ing a super majority vote by the board and/or the sharehold- The consulting-firm-led model lends itself to a ers attests to the importance attached to shared objectives governance discussion as it accounts for just over half of and “like-mindedness” in these projects. the cohort analyzed in this paper, and it provides a A management services contract (MSC) is often used to homogeneous structure for analysis across a number of structure and budget TA services from the holding compa- institutions. In consulting-firm-led greenfield MFIs, the nies, and/or the associated consulting firm, to the greenfield ownership structure includes the holding company, like- MFIs. Backstopping from the holding company and the tech- minded investors comprised mainly of DFIs and, in some nical services provided by the related consulting firm are cases, local shareholders, who appoint directors based on critical to the MFIs’ operating continuity, staff development, their level of ownership. Generally the approach has been and adherence to core operating policies and procedures. not to appoint executive directors; however, in some cases The board and shareholders play an important role in gov- regulatory requirements for resident directors have erning the potential conflict of interest that could otherwise necessitated the need, at least temporarily, for nonvoting arise from having staff from the consulting firm (acting on executive directorships. Independent directors are behalf of the holding company) involved in the decision- identified as management, and shareholders become more making process related to the procurement of technical ser- familiar with expertise that exists in the market. Board vices. Directors appointed by the holding company (which committees are gradually expanded as the MFI’s operations often is partly owned and managed by the consulting firm) become more complex. typically recuse themselves when the board votes on the 20 MSC. In other cases, decisions regarding the MSC are re- to the new greenfield MFIs. This was the case when Access ferred to the shareholders of the greenfield MFI, which typi- Holding launched a greenfield MFI in Liberia, for example. cally include several minority shareholders. Notwithstand- Institutional license. With only one exception, all of the ing the potential conflicts, the consulting firm’s ownership holding companies stress the importance of operating as an stake in the holding companies and, therefore, indirect own- institution that is permitted to mobilize deposits. Some of ership in the greenfield MFI, helps align financial interests. them have been willing to start with credit-only institutions but, when doing so, with the clear expectation to acquire a 3.3 Mission and Strategy deposit-taking license eventually. For these networks, depos- Without exception, the networks in the study are driven its are both a necessary service to customers and an impor- by a mission to expand access to financial services. They as- tant funding source for growth. This approach does not ap- pire to build retail mass market financial institutions that ply to BRAC, since its model in Africa focuses on delivery of serve populations neglected by mainstream banks, especially credit and social services. To secure funding for its greenfield micro, small, and medium-sized enterprises (MSMEs). Not MFIs, BRAC has created a debt-funding facility within its surprisingly, the extremely low-level of financial inclusion in group structure. SSA has been a key consideration in the networks’ decision The networks typically have two options to obtain depos- to enter and expand in the region. But they approach the it-taking capabilities in a particular country: a deposit-taking market in different ways, in terms of country selection, pre- microfinance license or a commercial bank license. Some of ferred institutional license (bank vs. NGO vs. nonbank finan- the consulting-firm-led networks have opted for commercial cial institution), and commercial orientation. bank licenses wherever possible, because they have the aspi- Country selection. The holding companies typically ration to create full-service banks capable of meeting most, if point to market characteristics, personal security of staff, po- not all, financial needs of MSMEs and low-income house- litical and economic stability, features of legal systems, and holds. However, minimum capital requirements for com- quality of financial sector supervision as significant factors in mercial banks are increasing across Africa, leading some net- determining country selection decisions. At the same time, works to opt for microfinance deposit-taking licenses with each of the holding companies has specific criteria for select- the intent to transition to commercial banks over time. Com- ing a market conducive to its unique implementation model. mercial bank licenses are, of course, subject to significant Most of the holding companies seek markets capable of sup- compliance cost and complexity, so not all networks are will- porting an efficient scale of operations and some of these ing to go that route when there are deposit-taking microfi- groups explicitly exclude smaller markets where they feel nance licenses available that can get the basic job done. that the impact and potential scale do not justify the invest- Commercial orientation. The holding companies are ment. On the other extreme, some holdings are taking a long- double-bottom-line investors, but they articulate their in- term approach to rural areas where they are less likely to vestment return expectations with different levels of preci- face competition. Most of the networks want a market that sion. Initially, the consulting-firm-led holding companies supports diversification into the SME segment. They con- expected to generate rates of return centered around 13–14 tend that the SME market is also underserved and that de- percent for their investors in a 6–8-year timeframe. The DFIs veloping SME products enables MFIs to maintain a partner- established this benchmark at the founding of the consult- ship with clients as they mature, allowing for continuity in ing-firm-led holding companies, and subsequent investors the financing relationship while protecting against attrition have entered with similar, if not higher, expectations. How- of the longest standing and most profitable relationships. ever, there have been few exits from the holding companies Finally, several groups deliberately seek a footprint that and it seems likely that the investment time horizon may be will allow for reasonably easy replication and sharing of ex- a few years longer than initially anticipated. It simply takes pertise and resources, for example, by focusing on a particu- more effort and more time than anticipated to build success- lar subregion. ASAI and BRAC are focused specifically on ful MFIs of significant size. English-speaking, common-law jurisdictions because it fa- As noted earlier, Fides has deliberately sought out inves- cilitates transfer of the model they have developed in Bangla- tors with a longer return horizon (10 years) and a more mod- desh. MicroCred is increasingly focusing on French-speak- est return expectation (8–10 percent) for SMH, which it feels ing West Africa, where the regulatory environment is is necessary when operating in the rural market. The NSO- relatively uniform. led holding companies speak in more general terms about While country selection is usually driven by the manage- expecting “reasonable” returns, and stress their primary ment and board of the holding companies, in some cases commitment to social objectives. DFIs and MIVs have made efforts to steer networks to cer- A common thread across all networks, regardless of their tain markets by promising significant and consistent support financial return expectations, is their aspiration to be so- 21 cially responsible. They all have a memorandum of associa- dardized set of systems, policies, and procedures: the IT plat- tion or articles of incorporation that define access to fi- form, internal control and risk management policies, and nance for micro and small entrepreneurs as a primary lending procedures, among others. Many networks have mission. Some networks, in particular BRAC, take this fur- gone through an extensive process of identifying, vetting, ther and attempt to provide or facilitate health and educa- and refining appropriate business practices for its MFIs. The tion services. Such services are generally linked to a credit- success of greenfield MFIs is in no small measure connected only approach to microfinance. to the effective implementation of such practices, and it is often cited as a critical success factor by the holding compa- 3.4 Institutional Capacity and Knowledge Transfer nies themselves. One holding company proudly calls itself The holding companies and sponsors typically take a pro- the “McDonalds” of microfinance, and contends that sys- active role in building institutional capacity among their tematic implementation of standardized systems is the key greenfield MFIs. They do so by providing technical and to cost and quality control. In particular, the adaptation and management services to the MFIs, which in turn are specified implementation of a common IT platform across the net- in a management services contract or a TA agreement. There work can yield significant advantages compared to stand- are some differences in how the networks have arranged the alone MFIs, in the form of better support services (provided delivery of these services. In the MicroCred network, for in part by the holding company or the sponsor), lower licens- example, the MFIs procure management services and TA ing costs, and greater ability to manage complex system re- directly from the holding company. In the SMH, Advans, and quirements. This is particularly important for greenfield Access Holding networks, MFIs procure much of this from MFIs, which have plenty of other challenges to worry about. the sponsors (Fides, Horus, and LFS, respectively). MFIs Almost all of the networks in this study have implemented or operating under BRAC’s Dutch holding company also procure are in the process of implementing a single core banking sys- services directly from the sponsor, BRAC Bangladesh. FINCA tem for their MFIs. MFIs obtain TA, oversight, and services from FINCA Training national staff. Holding companies identify hu- International and FMH in exchange for dividend payments. man resource development as the greatest challenge they The long-term vision for service delivery in the Ecobank face in SSA, more so than in other regions. The most com- model is less clear, as Ecobank has not yet developed mon approach is, like in other regions, to hire young adults comprehensive capacity to provide tailored expertise to its with little work experience and develop their skills and ca- network MFIs. Regardless of these differences, the knowledge pacity over time. Given that the average greenfield MFI has transfer from the holding company or the sponsor is usually about 300 employees after 30 months of operations, and channeled through three main mechanisms: (i) the more than 500 after 60 months, the training need is enor- international management team; (ii) the systems, policies, and mous. And many times the training must go back to the ba- procedures; and (iii) the training of national staff. sics, particularly in post-conflict countries. Often basic arith- Management team. Holding companies and/or sponsors metic skills need refreshing before staff can evaluate and usually provide anywhere from two to six managers for the present loan proposals. Most networks have relatively well- first three years of operations of a greenfield MFI. Typically established training modules for every key aspect of the busi- this includes the CEO, the chief operating officer, the bank- ness, facilitated by standard systems, policies, and proce- ing services manager (if the MFI takes deposits), and one or dures. It is also common practice for networks to move two temporary branch managers. These managers are meant employees among network MFIs as a way to build and trans- to provide skill and experience during the initial stage of fer knowledge within the group. operations. But they are also expected to provide internal ­ In addition to staff training that occurs at the MFIs and ­ coherence and a strong culture that reflect the values and staff transfers between the MFIs, many networks are set- mission of the network. They are as much role models and ting up regional hubs or regionally based staff that support advocates of corporate culture as they are technical experts. MFIs through technical expertise, backstopping, and man- This management team typically has a lot of authority in agement oversight. FINCA was one of the first to set up a guiding the delivery of specific technical services to the MFI: regional hub, which in addition to providing technical what services are needed, when they are brought in, and how backstopping, has direct management responsibility of they are implemented. Over time, most of these managers MFI CEOs, effectively decentralizing management. Simi- are replaced by national staff who have risen through the larly, but with a less robust approach, OI and Accion have ranks and proven their capacity to execute the required regional heads based in Africa. Other examples range from ­responsibilities. a training academy ProCredit set up in Ghana to the region- Systems, policies, and procedures. Another key advan- al decentralization of IT systems housed in Ghana for Ad- tage conferred by the holding company/sponsor is a stan- vans and in Senegal for MicroCred. (See Box 8.) 22 BOX 8 Cross-Fertilization in the Holding: Example of Advans Holding companies increasingly rotate national staff One of the internal trainers from Advans Cameroun coached among their network MFIs as a motivational measure, to client officers in Cote d’Ivoire. build a consistent group culture, and to share knowledge The exchange proved to be a win–win solution. Clearly on methods, standards, and procedures. Younger the arrangement helped Advans Cote d’Ivoire get off to a institutions benefit from expertise and culture built up in better start. But the staff exchanges also had a positive effect more mature greenfield MFIs at low or no added cost to on staff loyalty in Advans Cameroon, according to its CEO. them. The rotating staff remain either fully paid by the more Attrition rates have declined in recent years as staff recognize mature MFI or the host MFI takes on the direct costs of international growth opportunities within the network, which salary and travel. These rotations offer interesting growth are not offered by other MFIs or even local commercial banks opportunities for national staff, who start seeing the in Cameroun. possibility of an international career. In addition to staff exchanges, procedures and training One example is the assistance of Advans Cameroun, material are shared across the Advans network. The CEO of which was launched in 2007, to Advans Cote d’Ivoire, which Advans Cote d’Ivoire estimates that it saved two to three opened its first branch in March 2012. Advans Cote d’Ivoire months in international staff time compared to earlier greenfield relied extensively on support from Advans Cameroun to MFIs, because of materials available from Advans Cameroun. prepare and launch operations. Two Cameroonian branch Other benefits include pilots with new products and delivery managers worked in Cote d’Ivoire, and a group of Ivorian channels, which are tested in one MFI and then shared with loan officers got 4.5 months on-the-job training in Cameroun. others; this spreads the development costs over more entities. Source: Interviews with CEOs of Advans Cameroun and Cote d’Ivoire (December 2012). 3.5 Success Factors and Common Challenges clarity in organizational priorities. Clear organizational pri- Standardization of operational policies, procedures, and orities enable coherent long-term decision making for build- systems are the critical operational success factors most of- ing appropriate expertise and capacity at the holding com- ten cited by the holding companies themselves. They also pany, including human resources, IT system capabilities, cite close oversight and consistent reinforcement of internal tools and methods, and means of sharing knowledge. Capa- controls as critical to achieving a sustainable institutional ble networks tend to be guided and supported by a holding culture in greenfield MFIs. In the long term, the holding company that has a focused, long-term perspective, that en- companies contend that human resource development is key ables its management and staff to invest in the right exper- to success. tise, organize teams purposefully, and institutionalize effec- In addition to these factors, core strategic factors in- tive knowledge transfer practices. Holding companies or clude alignment of shareholder interests, clarity of vision sponsors that are not focused on creating and managing and mission, sponsor commitment, the regulatory and busi- MFIs are not very likely to be effective in greenfield MFIs, ness environment, and sufficient resources. When trans- given the difficulty and complexity of this business model. parent, these core strategic factors lead to success; how­ Sponsor commitment. As noted earlier, strong commit- ever, when muddled, they become significant challenges to ment can often come from a strong sense of mission. How- MFI sustainability. ever, to reinforce this commitment, holding company struc- Shareholder alignment. It is important that the share- tures typically aim to make sure that the sponsor and key holders involved, at the holding level and at the MFI level, decision makers and managers assume a financial stake in have a similar long-term vision for the network and its MFIs. the network. This becomes particularly important when If the long-term vision is shared and agreed to, most dis- things are not going as well as hoped. In those instances, the agreements will be about tactics and can be easily resolved. If MFI may not be able to pay for all services required, under- the long-term vision is not shared, or if it is unclear, funda- scoring the importance of having a sponsor and key staff that mental disagreements can arise and can require a lot of effort are willing to go the extra mile without necessarily being im- and energy to resolve—effort and energy that could be better mediately compensated for it. For example, some sponsors used to improve the MFIs’ operations. Some networks have will send additional junior consultants, or consultants with learned the hard way. strong technical skills but limited field experience, to sup- Clarity of vision and mission. Clarity of vision and mis- port the early operational stage of greenfield MFIs without sion that is centered on creating and managing MFIs leads to charging for it. In other cases, staff from mature network 23 MFIs are seconded on a subsidized or direct-cost basis to conditions. These problems are particularly acute in post- younger MFIs. Not only does this support the operations of conflict countries. the MFIs, but it also builds the experience and expertise of Sufficient resources. For many of the aforementioned rea- network and holding staff, which is likely to yield additional sons, the holding companies report that the cost of doing busi- returns to the network. ness in SSA is decidedly more expensive than the cost of their Enabling and predictable regulatory and business operations in other parts of the world. As a very simple exam- environment. Several networks identify the regulatory and ple, the cost of preparing a deposit-taking branch in Eastern supervisory regimes as a challenge to their activities. There Europe is estimated at $50,000 whereas in SSA it ranges from is broad frustration with protracted licensing procedures, $150,000 to $400,000. But other operating costs are also difficult communication with regulatory authorities, and a ­ higher, including communications, transport, and security. general lack of supervisory competence that, in some cases, The NSO-led holding companies, which have decades of has undermined the sectors’s credibility and stability. The experience in other parts of the world, also point to the chal- greenfield model discussed in this paper relies less on lenge of creating MFIs in SSA under ever increasing expecta- external supervision, since it is designed with very strong tions about financial performance. Available grant funding is internal controls and governance structures. more modest compared to when many Latin American MFIs In addition, greenfield MFIs in Africa face significant launched. At the same time, investors are more demanding challenges related to the business environment. Political about reaching breakeven in a three- to four-year timeframe. instability has been a constant threat and, in several cases, Several of the holding companies signaled some concern has imposed heavy costs on greenfield MFIs and holding about overly ambitious expectations from DFIs about finan- companies. Personal security is also a concern, as it adds to cial performance in a short timeframe. They see MFIs con- cost of operations and makes it more difficult to attract centrating in urban centers and a trend toward consumer international talent as necessary. Several networks also cite credit products that is elevating the danger of over indebted- the challenges of their international staff in adapting to local ness in some markets. 24 4 SECTION The Role of Greenfield MFIs in Market Development T he following section provides insights into market de- four greenfield MFIs served 89,942 microenterprise borrow- velopments associated with the start-up of greenfield ers and 265,714 depositors at the end of 2011, representing MFIs in the DRC, Ghana, and Madagascar.8 These approximately 50 percent of all borrowers in the microfi- three markets each have several greenfield MFIs and repre- nance sector and 65 percent of the depositors served by sent different country contexts in terms of financial sector MFIs.9 In that year, the microfinance sector counted 146 co- development. In each country at least two greenfield MFIs operatives and 16 local MFIs in addition to the four green- have been operational for more than five years, increasing field MFIs. When looking at the mainstream financial sector, the likelihood that effects of their interaction with the mar- the two greenfield MFIs with banking licenses represent a ket can be observed. This research shows that greenfield significant market share of loans and deposits. In 2011, Pro- MFIs play various roles in the development of the market for Credit DRC and Advans DRC accounted for more than 20 financial services for those at the bottom of the pyramid. In percent of all deposit accounts in the commercial banking addition to improving access to finance they also increase sector and 13 percent of all commercial bank loans. When the level of skills in the financial sector, introduce new prod- adding loan accounts from the other greenfield MFIs, the ucts and channels to the market, and expand the number of market share increases to 61 percent of all loans in the finan- access points for clients. cial sector.10 Their deposit volume and loan portfolio repre- The analysis in these markets is largely based on inter- sented about 7 percent of the financial sector. However, views with stakeholders, direct data collection from the more remains to be achieved as the penetration rate remains greenfield MFIs, and secondary data sources for the microfi- low overall with only 5 percent of the Congolese adult popu- nance and broader financial sector in each country. While it lation banked.11 is difficult to attribute changes in a market comprised of In Madagascar AccèsBanque and MicroCred together many institutions to the intervention of one or more green- held US$33 million in deposits and US$35 million in loan field MFIs, the authors used mostly qualitative and some portfolio at the end of 2011. This represented only 2 percent quantitative information to examine the effect they have had. and 5 percent of the banking sector, respectively, but in num- Throughout the text, it is indicated where the anecdotal dis- ber of deposit and loan accounts the two greenfield MFIs ac- covery method supports assertions, but also where this may counted for 25 percent and 17 percent, respectively, of the 11 not be the case. banks in the system. When compared to the 36 other MFIs in the country12 the two greenfield MFIs held almost half of the 4.1 Market Relevance entire microcredit portfolio and close to a quarter of MFI de- The greenfield MFIs in the DRC, Ghana, and Madagascar posit balances. AccèsBanque and MicroCred have clearly represent only a small portion of total financial sector assets, contributed to the rapid growth of microfinance in Madagas- but they are significant players in terms of numbers of house- car, which reached a penetration rate of 21 percent in 2012 holds and enterprises served. In some cases, they also man- compared to 14 percent in 2008. The banking sector (exclud- age a significant number of branches and employ a signifi- ing the two greenfield MFIs) reached only 3 percent of the cant number of employees relative to the financial sector population in 2012 (up from 1.8 percent in 2008). overall. See Table 9. In Ghana, a more advanced, competitive, and diverse fi- These effects are observed most clearly in countries with nancial sector compared to DRC and Madagascar, green- a less developed financial sector. In post-conflict DRC, the field MFIs represented five of the 19 savings and loans com- 8. G reenfields included in the analysis were in DRC: Advans (2009), FINCA (2003), ProCredit (2005), and OI (2010); in Ghana: EB Accion (2008), Advans Ghana (2008), ProCredit (2002), OI (2004), and ASA (2008); and in Madagascar: ABM (2007) and MicroCred (2006). Market analysis has been per- formed by Yaw Brantuo for Ghana, Hannah Siedek for DRC, and FTHM Conseils for Madagascar. Fund for the Financial Inclusion in DRC (FPM), Synthesis of the political, economic, and financial evolution of DRC. Facility for financial inclusion in 9.  DCR. Third trimester 2012. Kinshasa, DRC. 10. Data based on the International Monetary Fund (IMF) Financial Access Survey (FAS) database, which includes information for commercial banks only. 11. FINDEX, http://siteresources.worldbank.org/EXTGLOBALFIN/Resources/8519638-1332259343991/N4ssaEN_08202012.pdf, accessed January 2013. 12. For this market analysis the two greenfield MFIs in Madagascar are compared to the 36 MFIs (five “etablissements financier” and 31 “institutions de microfinance agrees”) as their clientele, and methodology is more aligned than with the other nine commercial banks in the country. 25 TABLE 9 Market Share of Greenfield MFIs in DRC, Ghana, and Madagascar, 2011 Deposits Loans Total deposit volume Number of Gross loan portfolio Number of (USD) deposit accounts (USD) loan accounts DRC Greenfield MFIs 141,917,898 317,217 71,005,502 93,640 % of commercial banks and Greenfield MFIs 7.1 21.2 6.7 60.9 Ghana Greenfield MFIs 82,619,026 514,258 90,424,087 149,669 % of commercial banks and Greenfield MFIs 0.9 7.2 1.5 21.7 Madagascar Greenfield MFIs 33,351,084 133,538 35,276,058 37,072 % of commercial banks and Greenfield MFIs 23.4 25.2 4.8 17.3 Notes: For the DRC and Ghana, the FAS database includes data for commercial banks only and not data for other financial institutions such as those in Madagascar. To calculate market shares, the data of greenfield MFIs that are not included in FAS have been added to commercial bank data to establish the denominator (FINCA and OI in the DRC and all greenfield MFIs for Ghana). For Ghana the share of loan accounts for greenfield MFIs could be somewhat overestimated as FAS data included number of borrowers for commercial banks only, which tends to be lower than the number of loan accounts. Sources: Greenfield MFIs and IMF FAS database, 2011. panies.13 Together the greenfield MFIs served more than outstanding loan portfolio is modest at 1.5 percent of the 150,000 borrowers and over 540,000 depositors at the end financial sector while their loan accounts represent a sub- of 2011, which represents around 5.1 percent of all adults or stantial 22 percent (see Figure 9). 2.8 percent of the entire Ghanaian population (see Figure 8).14 Their deposit volumes remain modest as a percentage 4.2 Skills Building of the financial sector (in this case, defined as commercial By their own account, the greenfield MFIs’ most signifi- banks plus savings and loan companies) at 0.9 percent but cant effect on market development is through their contribu- they hold 7 percent of deposit accounts. Similarly, their tion to the professional development of staff in the banking FIGURE 8 FIGURE 9 Number of Deposit Accounts, Number of Total Loans Outstanding, December 2011 December 2011 8,000,000 800,000 7,000,000 700,000 6,000,000 600,000 5,000,000 500,000 4,000,000 400,000 3,000,000 300,000 2,000,000 200,000 1,000,000 100,000 0 Madagascar Ghana DRC 0 Madagascar Ghana DRC Green eld MFIs Commercial banks excl green eld MFIs Green eld MFIs Commercial banks excl green eld MFIs At the time of this study, ASA was in the process of transforming from an NGO to savings and loan. For the purpose of the study it has been included 13.  among the greenfields, which together are compared with industry data for deposit-taking banks in Ghana (including commercial banks, rural banks, and other banking and quasi-banking institutions). To put this in perspective, in 2005, all mainstream commercial banks together had an estimated penetration of 5 percent of the overall population. No 14.  more current comparative information was available. 26 and microfinance sectors. According to market participants, the greenfield MFIs have introduced superior human re- BOX 9 source practices that positively impact the financial sector. ProCredit Young Bankers Program With the exception of a small number of international staff, all 11,600 employed in greenfield MFIs as of December 2012 ProCredit’s Young Bankers Program is an intense, are nationals. The number of people employed and trained half-year training on banking and finance for university by the greenfield MFIs is therefore becoming significant in graduates with little or no practical work experience. relation to the overall financial sector in many countries. In ProCredit reported that its recruiters for the program Ghana, greenfield MFIs employed more than 2,000 staff in look for individuals with good analytical, organizational, 2011 while the mainstream banking sector employed 16,000 and communication skills who are capable of solid staff. The two greenfield MFIs in Madagascar have more quantitative analysis, have demonstrated ability to think than 1,000 staff, which represents 23 percent of staff in the logically and critically, and can work effectively in teams. microfinance sector and almost 19 percent of banking sector Such individuals are also required to show a clear desire employees. The employees of greenfield MFIs—typically and ambition to learn and develop in the profession. young adults who have little or no previous work experi- Those selected to participate in the program are taught basic mathematics, accounting, and other relevant ence—receive extensive training in several topics and skills banking subjects. But more importantly, the training related to credit and banking. Eventually they become at- emphasizes both theoretical and practical training with a tractive candidates for mainstream banks, and their skills view to instilling good banking practices and the gradually are incorporated into the larger market as their ca- ProCredit culture and methodologies for doing business. reers bring them to other institutions. The positive results to This sort of investment in staff training is complemented, the financial sector from the large investments in staff train- for example, by advanced training in ProCredit’s regional ing and development by greenfield MFIs reduces the poten- training center for middle managers in Macedonia and tial market distortion from providing TA grant funding to its international training center in Germany, which individual institutions. provides a three-year part-time training course for senior Greenfield MFIs typically have an intensive and systematic managers. approach to staff selection, recruitment, and training. They Source: Yaw Brantuo from interviews and ProCredit website [www.procredit-holding.com] spend 3–5 percent of their operating budget on staff development. This is where a significant portion of the initial TA resources is invested. Most greenfield MFIs have fered by greenfield MFIs in Madagascar appear to be ­higher16 company-specific training facilities that offer courses for than the average for the microfinance and banking sectors. induction and professional development. They also offer According to those interviewed, this has had a notable effect intensive on-the-job training. Many of the people interviewed on some of the MFIs in Madagascar, especially ACEP and in the DRC, Ghana, and Madagascar commented on the high BNI-CL, which have seen staff leave for better opportunities quality of the training offered by greenfield MFIs. at the two greenfield MFIs. The greenfield MFIs appear to Mainstream banks and other financial institutions appear provide a career bridge between the less formal microfi- to agree because they frequently try to poach staff from nance sector and the more formal banking sector. greenfield MFIs. Some holding companies calculate that they will train two to three times the number of required 4.3 Product and Channel Diversification staff to address expected attrition to local financial institu- Greenfield MFIs tend to be at the forefront (compared to tions. Staff turnover rates reported by the greenfield MFIs other MFIs) of introducing innovation in low-income retail varied between 8 percent and 20 percent.15 For example, Fi- banking. Greenfield MFIs have introduced new products, delity Bank Ghana hired staff of ProCredit to support its credit policies, and service standards that have been replicated branch roll out. Several of the banks that entered the SME by other financial institutions. For example, in the DRC, Pro- space in the DRC are also run by former ProCredit staff. (See Credit introduced free savings accounts without a minimum Box 9.) deposit requirement at a time when most banks had minimum Staff compensation varies across markets. The greenfield requirements of more than US$1,000. ProCredit attracted MFIs in the DRC tend to offer less attractive packages than large numbers of savers and demonstrated that the Congolese mainstream banks and some MFIs. In contrast, salaries of- population was able and willing to save. Following this exam- This compares with turnover rates of only 2–3 percent for other MFIs reporting to MIX in the DRC and Madagascar. In Ghana, turnover rates appear to 15.  be high (10–20 percent) for most MFIs that reported these data to MIX. In Madagascar this is reported to be 20–100 percent higher, depending on the function. For example, a loan officer with two years of experience and a 16.  baccalaureate received an average salary of US$120 in an MFI, US$150 in a mainstream bank, and US$210 in a greenfield MFI. 27 ple, other banks, such as Rawbank and BIAC, relaxed their ricultural finance. OI started an agricultural finance program account-opening requirements, and the number of deposit ac- in Ghana in 2010 with a pilot credit scheme for cocoa ­farmers. counts in the DRC has grown from 30,000 in 2005 to 1 million It now serves 9,000 farmers and has introduced geographic in 2012. Similarly, Malagasy MFIs adapted their internal pro- information system (GIS) technology to more accurately cedures, processes, and IT systems to keep up with the new map the smallholder farmers (see Box 10). greenfield competition, evidenced by the reduction in loan In a few cases, the success in SME lending of greenfield processing times from weeks to five days. MFIs has attracted other providers into this market segment. In Ghana and the DRC, greenfield MFIs were the first to According to observers in the DRC, the performance of Pro- introduce new technologies in banking for low-income pop- Credit, which reached financial sustainability in three years, ulations. EB-Accion, Opportunity, and ProCredit introduced triggered banks such as BIC and TMB to downscale and ATMs in Ghana, which were previously available only at serve the SME segment. To acquire the necessary expertise, commercial banks. EB-Accion Ghana and Advans Ghana, for they relied on former employees from ProCredit to roll out their part, have introduced mobile deposit collection (using these services. In Madagascar, Bank of Africa (a shareholder cell phones for instant verification) to offer clients additional of MicroCred) and BFV-Societe Generale (a shareholder of convenience and to compete with traditional “susu” collec- AccèsBanque Madagascar) have started to modestly down- tors. In the DRC, ProCredit established the first ATMs, and scale in the past two years. Whereas five years ago Malagasy mainstream banks soon followed. Clients in the DRC now SMEs were not served at all, there is now an increased offer have access to point-of-sale (POS) devices at over 300 loca- of services from a range of financial institutions. tions, facilitating the withdrawal of funds and cashless ­purchases. 4.4 Standards and Good Practices Some greenfield MFIs have pioneered the development of Greenfield MFIs can play an important role in market de- financial services that are perceived as particularly risky and velopment by demonstrating professionalism and good prac- challenging in their markets, such as microinsurance and ag- tices. Evidence of this is not easy to establish but some signs BOX 10 Greenfield Banks Pioneering with Smallholder Finance Smallholder finance is typically avoided by mainstream methodology is now in the process of being transferred to banks and considered too risky. Several of the more mature AccessBank Tanzania. greenfield MFIs, such as ABM in Madagascar and OI in OI Ghana piloted an agricultural finance program in Ghana, have piloted agricultural finance products to 2010 with an input credit scheme using 536 cocoa farmers address this gap. in the Ashanti region. By 2012 OI Ghana was serving about After four years in operation, AccessBanque Madagascar 9,000 farmers across six administrative regions of Ghana. launched an agricultural loan product “Agro loan” for OI Ghana is introducing the use of GIS technology to more smallholder farmers who mix subsistence farming with accurately map the locations of smallholder farmers to some production of cash crops. ABM uses a cash-flow- enable them to accurately determine the areas under based methodology and adapts its loan terms to the cultivation and avoid over or under use of agro-inputs, farmers’ needs with variable monthly payments (according which has negative impact on the yield. In Ghana, where to client’s cash flow) and a principal grace period, if needed. agricultural finance is only 6 percent of the loan portfolio of The pricing of the loan is similar to a microenterprise loan. the entire banking sector and the quality of these loans are By 2012, ABM had extended 1,700 agro loans, which below industry average, a breakthrough in this area could represented around 7 percent of its clients and 2.4 percent impact the sector. of the loan portfolio. The average loan size was around While AMB and OI’s experiences are still at a small scale, US$419. ABM offers this product in a radius of 25 km around their performance is watched carefully by others. the branches because loan supervision requires close Experiences such as these of ABM and OI could potentially monitoring. ABM has a dedicated team of agricultural loan provide results that would widen the overall contribution of officers that receives in-house and external training on the the banking sector to the domestic economy and positively topic and receives slightly higher compensation than the impact the lives of the rural population, which is still severely enterprise loan officers because of their specialization. This underserved. Sources: Agrifin (2012) and Yaw Brantuo. 28 are nevertheless visible. Greenfield MFIs generally apply MFIs19 to have potential access to financial services from high standards related to transparency with clients and are entities in both sectors. also often active contributors to national credit reference bu- Finally, some greenfield MFIs contribute to the develop- reaus. Also, greenfield MFIs have on several occasions advo- ment of market infrastructure and a more favorable regula- cated changes on behalf of the microfinance sector, to en- tory environment by participating in the banking or MFI as- hance transparency, raise standards, and improve the quality sociations in their country. ProCredit in the DRC was of regulations. See Box 11. instrumental in negotiating a liquidity ratio in favor of the Several holdings endorse the Client Protection Princi- microfinance sector with the Central Bank.20 Advans Camer- ples,17 and train their staff to operationalize these principles. oun contributed training material on know-your-customer In the DRC, Advans and ProCredit led the way in more trans- requirements to the Central Bank for a workshop provided parency toward their clients, and now Rawbank and BIC, to all MFIs in the country. In Ghana, greenfield MFIs played two traditional commercial banks, also publish their prices a role in lobbying for measures to allow savings and loans and terms on their websites. In Ghana, market actors inter- companies to clear checks and engage in foreign currency viewed found greenfield banks to be more open and trans- denominated transactions. parent in their dealings with their clients. This view is sup- ported by the availability of client-oriented material on the BOX 11 websites of greenfield MFIs and the clearly visible pricing information posted in the banks. In Madagascar, Accès- ASA International, an Impact Investment Banque Madagascar is one of only two MFIs that publish ef- Pioneer in Transparency fective interest rates for their clients. Greenfield banks comply with reporting to the credit ASA International is among the first funds and holding companies to be rated by the Global Impact bureau or are, depending on the country context, actively Investment Rating Service (GIIRS). The rating provides participating in exchange of credit references among an independent judgment of social and environmental institutions operating in the same market. The holding impact, practices, policies, and achievement and covers companies highlight the importance of this for sound credit the holding, investment fund (CMI), as well as the risk management and responsible finance.18 In Cameroun, underlying investments, ASAI’s greenfield MFIs in Asia for example, where Advans built its first greenfield MFI, the and Africa. While the rating is preliminary and not public, bank started a credit information exchange that now involves it provides ASAI a benchmark relative to other impact 12 MFIs. In Ghana greenfield MFIs are all active providers investing companies. In 2012 ASAI received a rating of and users of credit reference information, as required by law. 161 points out of a total of 200, which ranks it among In Madagascar, greenfield MFIs (which are licensed as the top quintile of the early GIIRS-rated companies, the commercial banks) willingly report to two credit bureaus: so-called Pioneers. ASAI envisages to regularly update one for banks and one for MFIs. This dual reporting creates its rating and is currently undergoing a follow-up rating. a bridge between the larger microfinance sector and the Source: GIIRS (2012) and GIIRS website (www.giirs.org). banking sector, enabling the clients of these two greenfield  he Smart Campaign website (www.smartcampaign.org) lists Access, Accion, Advans, BRAC, FINCA, MicroCred, OI, and Swiss Microfinance holding 17. T as endorsers as well as some of their individual affiliates in SSA. 18. In Ghana XDS Data Ghana noted that the greenfield MFIs were among the few financial institutions that undertook direct and extensive due diligence of the credit reference system before signing on. Executives of greenfield MFIs undertook onsite visits to observe reliability, safety, and adequacy of the equipment and related processes of their facilities. 19. Other MFIs do not have access to the bank credit bureau. 20. The liquidity regulation in the DRC was defined in such a way that current and savings accounts needed to be covered by the same liquidity even though savings accounts were four times less liquid. ProCredit DRC demonstrated this with actual data in discussions with the Central Bank following which the ratio was revised to 60 percent for current accounts and 40 percent for savings accounts. 29 30 5 SECTION Conclusion T he greenfield model has come a long way in a short Where does the model go from here? The number of time in SSA. While there is a range of microfinance greenfield MFIs has grown rapidly in the past decade, and providers in SSA, the proliferation of greenfield MFIs new entities are still being added to this segment of the mi- expands the commercial end of the spectrum with regulated, crofinance industry. It is likely, however, that the rate of cre- deposit-taking institutions focused on microenterprises and ation of greenfield entities, at least in SSA, will slow, as the small businesses. In most countries there are gaps in every most “feasible” markets have now largely been entered. But segment of the market. Many greenfield MFIs address the this leaves about 25 countries in SSA without any greenfield broader micro and small business segments, while many ex- MFI presence, and typically without the presence of any sus- isting MFIs tend to cater to microenterprise alone. Holding tainable MFIs at all. Some have populations too small to sus- companies have established promising institutions in very tain a commercial institution (many island nations), some difficult markets, including post-conflict markets such as the are unstable and in conflict (Somalia), some have a combina- DRC, Cote d’Ivoire, and Liberia. While many greenfield tion of these issues (the Central African Republic), and some MFIs are still young, there are signs of solid institution build- have limitations on foreign ownership in the banking sector ing for the longer term and positive effects on local markets. (Ethiopia). And in almost every country, peri-urban and ru- Holding companies and their shareholders have found ral populations still struggle to access financial services. ways to leverage the greenfield business model to address One challenge for greenfield MFIs and their holding com- some of the primary challenges in advancing access to fi- panies is, therefore, to develop a delivery model that facili- nance in SSA. The small size of nascent financial markets, tates commercially viable and affordable access in smaller, high costs of doing business, uneven regulatory frameworks, more dispersed markets and rural areas. Indeed, some of the and inadequately skilled human resources in many SSA mar- more mature greenfield MFIs that have achieved breakeven kets benefit from an approach where practices can be stan- are now exploring alternative delivery channels, such as dardized and costs can be shared. Through network struc- agent banking and mobile financial services, to extend their tures and common practices, holding companies are able to reach in markets with low population densities that present transfer knowledge and learning from one greenfield MFI to challenges for traditional bricks-and-mortar expansion another, leveraging the investment in human resources and models. Likewise, their product development is pushing the skills across borders. Their structured approach and heavy boundaries at both ends of the spectrum with a focus on focus on human resources development has been a key suc- SME lending at one end and payment, credit, and deposit cess factor in their ability to create sustainable institutions in services for the mass market at the other. some of the most frontier markets in SSA. Greenfield MFIs Another possible development is that some holding com- have kept risks and losses relatively low thanks to rigorous panies will develop greater appetite and capacity to acquire training of staff, consistent application of tested methodolo- existing MFIs and small business banks. So far this has been gies, and strong commitment to strict quality and service a rare occurrence indeed, as the holding companies are usu- standards. ally apprehensive about the costs and risks involved in the The sustainable performance of greenfield MFIs illus- acquisition of other entities. However, it is possible that a trates to the traditional formal banking sector that under- dearth of attractive greenfield opportunities will create in- served businesses and households are bankable, and even centives for some holding companies to consider such op- profitable, market segments. Market analysis in the DRC, tions, particularly if they feel they have a strong network of Ghana, and Madagascar has shown that greenfield MFI MFIs that could support the operational aspects of such an practices have often been transferred to other market par- approach. ticipants by example and through staff movement. Even At the same time as the holding companies and greenfield though attribution remains a difficult issue, it appears, espe- MFIs face significant operational challenges (and cially in less-developed financial markets, that greenfield opportunities), they will also have to manage their investors’ MFIs play a pioneering role in expanding the financial access expectations, particularly those of the DFIs. Proof of of microenterprises, small businesses, and low-income concept now has to give way to mass market reach and households. shareholder returns. Apart from the pressure that this 31 creates, it will also generate questions about the continued could occur as a result of an IPO at the holding company role of DFIs in greenfield networks, with some arguing that level, at which point many DFIs may exit their investments there is a continued role (e.g., by helping achieve mass in both the holding company and greenfield MFIs. market reach) and others arguing that the DFI role has The greenfield MFI model is a complement to other largely been fulfilled and that there is time to realize some strategies for increasing access to finance, such as reform of financial returns. existing institutions without a holding structure and bank The shareholding of greenfield MFIs has been very downscaling. The next few years will be very telling about stable so far, but it is possible that the market will see more the ability of greenfield MFIs to leverage their foundation movement in the ownership of these entities going forward, and achieve scale, and for holding companies to replicate particularly if local investors start taking a greater interest. and sustain the success of their model in other markets, It is also possible, though not very likely, that the market particularly in a context of diminishing funding for TA. could see sales of entire greenfield entities if they are not Undoubtedly they will find themselves compelled to develop deemed to fit the future strategy of the network. In some new methods, capacities, and practices to stay relevant and cases, DFIs have pursued a dual-stage exit, swapping shares competitive in the microfinance space. This is already from the local subsidiaries into the holding. The evident in the increasing level of investment directed toward diversification provided by a holding company portfolio technology-based solutions and alternative delivery presumably makes the investment more liquid and channels. At the same time, it is also likely that many of these potentially easier to exit through a sale of shares to new greenfield MFIs will increasingly begin to compete with holding investors or the potential exit through an initial commercial banks for mass market customers and those in public offering (IPO) in the capital markets. Finally, there is the SME space. The financial landscape in Africa is poised to the (even more remote) possibility that a major reshuffle become much more interesting. 32 1 ANNEX Methodology Definition of Greenfield MFI mained credit-only MFIs, mobilizing compulsory savings, Greenfield MFIs are defined for the purpose of this but not taking voluntary deposits from clients other than publication as institutions that are newly created without their members. pre-existing infrastructure, staff, clients, or portfolios and Performance of the greenfield MFIs is evaluated by insti- use standard operating procedures disseminated by a central tutional age (as opposed to calendar year) to evaluate prog- group, often a holding company. The holding company ress and maturation from start-up, regardless of the year op- usually also plays a strong role in backstopping operations, erations were initiated. This way, conclusions can be drawn providing standard policies and procedures, and co-branding at different stages of institutional development: foundation the subsidiaries in the network. (preparation and first year of operations), institutional de- Takeovers and mergers were not included, nor were some velopment (generally year two through breakeven), and subsidiaries of OI and FINCA that started as NGOs many scale-up (from financial breakeven onward). Among the 30 years ago and did not have data available from the start of institutions in the sample, five are in the foundational stage, their operations. Also, at that point in time, they weren’t part 11 in institutional development stage, and 14 in scale-up of a network governed by a holding company. stage. The number of greenfield MFIs in the sample for the Ecobank is the only commercial African bank included in different performance indicators and ratios therefore gets the sample. Some other large commercial banks, such as Eq- smaller closer to the month 60 timeline. uity Bank and UBA, which also provide microfinance ser- Performance data are presented as simple averages unless vices, have expanded to other countries. However, these in- otherwise indicated to display the performance of a typical stitutions do not focus exclusively on the MSE market and greenfield MFI. When outliers appeared to distort the sim- have not created a separate structure to distinguish their mi- ple average, weighted averages have been used (e.g., in the crofinance operations as Ecobank has made it more compli- case of return on assets and return on equity). cated to separate comparable date. Therefore, these banking groups have not been included in the study. Performance Benchmark with MIX Young Africa In Section 1 the performance of greenfield MFIs has been Data from African Greenfield MFIs and Life Cycle benchmarked to the MIX index for “young” MFIs in Africa, Performance Analysis which comprises 58 institutions, four to seven years old. All Eleven holding companies provided detailed performance but five of these institutions are deposit taking, with a de- data throughout the life cycle for 30 of the 33 African posit base ranging from $30,000 to $22.6 million ($2 million greenfield MFIs created in SSA since 2000. ProCredit on average) and gross loan portfolio ranging from $2,000 to Angola, Ghana, and Mozambique are not included in this $24 million ($2.7 million on average). The greenfield MFI detailed analysis as ProCredit Holding did not provide subjects of this paper were removed from the benchmark performance data for these banks by institutional age. The population. holding allowed the authors to use data available for two of their banks (DRC and Sierra Leone) that were readily Market Share Analysis of the DRC, Ghana, and available. ProCredit Ghana was included in the market Madagascar development research (see below). Data for ProCredit Sierra Greenfield MFIs that were included in the market level Leone are included only for the period in which it was owned analysis in the DRC, Ghana, and Madagascar presented in by ProCredit, from 2007 until 2010, when it was sold to Section 4 are (with their year of creation): DRC: Advans ­Ecobank. (2009), FINCA (2003), ProCredit (2005), and OI (2010); in The 30 greenfield MFIs included in the research repre- Ghana: EB Accion (2008), Advans Ghana (2008), ProCredit sent various types of institutions. A majority are licensed and (2002), OI (2004), and ASA (2008); and in Madagascar: ABM regulated deposit-taking institutions, ranging in legal struc- (2007) and MicroCred (2006). ture from commercial banks to savings-and-loan companies To determine the market share of these greenfield MFIs— to specialized deposit-taking MFIs. A few started taking de- in terms of deposits, gross loan portfolio, number of deposi- posits a couple of years after their creation, and a few re- tors, and number of loan clients—the greenfield MFI data 33 collected for this study were compared to data available in greenfield MFIs. Their data have been added to commercial the IMF FAS database 2011 data. In addition, some compari- banks data to establish the denominator for the calculation sons were made with the microfinance sector or financial of the market shares. For Ghana the number of loan accounts sector as a whole, depending on data availability. is likely overestimated as FAS data included only number of borrowers, which tends to be lower than the number of loan DRC ­accounts. For the DRC the FAS database includes data for commer- In Ghana the greenfield MFIs are five of the 19 savings- cial banks only and not for the rest of the financial institu- and-loans companies that represent only a very small part of tions. To calculate market shares, data of greenfield MFIs the financial sector. At the time of the study ASA in Ghana that are not included in FAS (i.e., FINCA and OI) have been was in the process of transforming from an NGO to a savings- added to commercial bank data to establish the denominator. and-loan company. In Ghana, data for the greenfield banks For deposit accounts, FAS contained data for all deposit-tak- are compared with industry data for so-called deposit money ing institutions, so this market share reflects the share of all banks, including commercial banks, rural banks, and other greenfield MFIs of the entire financial sector. banking and quasi-banking institutions. For the comparison with other MFIs, the market share analysis relies on data collected by the Fonds pour l’inclusion Madagascar financiere en RD Congo (FPM). FPM reported 146 coopera- The FAS database has disaggregated figures for Madagas- tives, 16 local MFIs, and six international MFIs and microfi- car facilitating the calculation of the market shares. Both nance banks (including the greenfield MFIs in the sample) greenfield MFIs are commercial banks. However, for the and three universal banks providing microfinance. purposes of the market analysis, they are compared with the 36 Malagasy MFIs (five “etablissements financier” and 31 in- Ghana stitutions de microfinance agrees”) as their clientele and Also for Ghana the FAS database includes data only for methodology are more aligned than with the other nine commercial banks data so no data were available for commercial banks in the country. 34 2 ANNEX List of people interviewed Holdings Opportunity International DRC, Gilbert Lagaillarde, Access Holding, Thomas Engelhardt, Management Directeur Generale Board Member Opportunity International Ghana, Kwame Owusu – Access Holding, Christoph Diehl, Management Board Boateng, Deputy/Acting CEO Member Accion International, Brian Kuwik, Regional Head Africa DRC Accion Investments, John Fisher, Vice President Jean Claude Thetika, Directeur General, Fonds pour Advans, Claude Falgon, CEO l’inclusion financiere en RD Congo (FPM) AKAM, Mwaghazi Mwachofi, CEO Michel Losembe, Directeur General BAnque ASA International Holding, Martijn Bollen, General Internationale pour l`Afrique Au Congo, President of Counsel ACB (Association Congolaise des Banques) ASA International Holding, Mischa Assink, Senior Accountant Ghana BRAC, Ishtiaq Mohiuddin, Director Microfinance Philip Cobbinah, Deputy Head, Banking Supervision, BRAC, Tanwir Rahman, Director of Finance Bank of Ghana FINCA, Helen Lin, Africa Finance Manager William Asare, Bank Examiner, Banking Supervision FINCA, Mike Gama-Lobo, Vice President and Regional Department, Bank of Ghana Director Africa Yaw Gyima-Larbi, Head, Microfinance Unit, Banking MicroCred, Arnaud Ventura, CEO Supervision Depart, Bank of Ghana Opportunity International, Colin McCormack, Head of Gloria Quartey, Head, Center for Training and Africa Operations Professional Development, Bank of Ghana Opportunity International, Jean-Philippe Nefve, CFO of Yvonne Quansah, Director, Financial Institutions Sector, Global Microfinance Operations Ministry of Finance EcoBank, Francis Adu-Mante, Managing Director Bernard Joe Appeah, Principal Consultant, Pentax EB-Accion Ghana Management Consulting ProCredit Holding, Helen Alexander, Management Team Emmanuel Owusu, Managing Director/President of Member Association, Global Access Savings and Loans/ Ghana Swiss Microfinance Holding, Thi Hanh, Operations Association of Savings and Loans Companies Manager FIDES Yaw Gyamfi, Executive Secretary, Ghana Network of Swiss Microfinance Holding, Christian Baron Microfinance Companies Swiss Microfinance Holding, Konrad Ellsasser Raymond Mensah, M&E Specialist, Rural and Agricultural Finance Programme (IFAD/Ministry of Affiliates Finance) AccesBanque Madagascar, Philip Acton, Richard Amaning, Executive Secretary, Ghana AccessBank Tanzania, Roland Coulon, CEO, Association of Microfinance Companies Advans Cameroun, Frank Snieders, CEO Vera Geraldo-Stephenson, Assistant Sales Manager, XDS Advans Cote d’Ivoire, Gregoire Danel-Fedou, CEO Data Ghana Limited Advans Bank DRC, Francois Lecuyer, Directeur Generale Advans Ghana, Cedric Henot, CEO Madagascar Advans Tanzania, Peter Moelders, CEO Antoine Rakotondrasoalimangarivelo, Directeur des ASA Ghana, Mohammed Aourongjeb, Executive Director Portefeuilles, TITEM EB-Accion, Ghana, Frances Adu-Mante, Managing Bakoly T. Rafanoharana, Expert National, PAFIM Director Blaise Francis Rajoelina, Coordonnateur National de la FINCA DRC, Ed Greenwood, CEO Microfinance, CNM MicroCred Madagascar, Barnabe Francois, CEO Brillant Rakotoarison, Directeur Général, SIPEM Sa. 35 Charlot Razakaharivelo, Directeur Général, FIDEV Randrianiaina Rakotoarivao, Directeur du Réseau, Fanjaharivola Rakotomaharo, Secrétaire Général, OTIV TANA APIMF Thomas Rasolonjatovo, Président Conseil Jean Herley Ambinitsoarivelo, Directeur, CEFOR d’Administration, MECI José Serge Rajaonarison, Directeur Général, Youssouf Mahamoud, Directeur des Opérations, OTIV CECAM DIANA Jules Théodore Rakotondramanga, Secrétaire Général, CSBF Investment officers Liva Claude Herimanana, Directeur Général Adjoint, AFDB, Robert Zegers, Rafael Jabba, Barnett Douglas and ACEP Timo Teinila Mahefa Edouard Randriamiarisoa, Directeur Général, FMO, Andrew Shaw, and Maurice Scheepens ACEP IFC, Adam Sorensen Ndriana Ralaimanisa, Directeur Commercial et KfW, Matthias Adler, Monika Beck, Simon Bleidiesel, and Marketing, BNI-CL Karl-Heinz Fleischhacker 36 3 ANNEX Bibliography AccesBanque. 2012. Financial Statements 2008–2012. EB-ACCION. 2013. EB-ACCION 2013 Status Report. Accra: Madagascar: AccesBanque. EB-ACCION. Advans Ghana. 2012. Advans Ghana Competitor Analysis. FinMark Trust. 2010. Finscope Ghana Survey 2010 Report. Accra: Advans Ghana. Accra: Finscope. Agrifin. 2012. “Banking Madagascar’s Small Farmers: FPM. 2012. “Synthesis of the Political, Economic and ABM’s Cash Flow-Based Agricultural Credit Analysis Financial Evolution of DRC. Facility for Financial Methodology.” Washington, D.C.: Agrifin. Inclusion in DRC. ” Third trimester 2012. Kinshasa : Ahouissou, B., and M. Andriamahenina. 2012 “Final DRC. Evaluation Mission Report of the National Microfinance Fritz, B., and K. Hujo. 2005. Economy under Latin American Strategy, SNMF 2008–2012.” Antananarivo: UNDP and Conditions. Frankfurt am Main: Vervuert, pp. 95–122. UNCDF. https://www.lamicrofinance.org/files/34227_ Gaul, Scott. 2011. “Mapping Africa Financial Inclusion— file_evaluation_SNMF_Madagascar.pdf Results Review and Next Steps.” Washington, D.C.: MIX. Assouline, N., and C. Poursat. 2010. “Malagasy MFIs Facing h t t p : // w w w. t h e m i x . o r g / p u b l i c a t i o n s / m i x - the Crises: What Lessons to Learn?” AFD. microfinance -world/2011/09/africa-financial- Bank of Ghana. 2010. Bank of Ghana Annual Report 2010. inclusion/results#ixzz2pNipbrQi Accra: Bank of Ghana. GIIRS. 2012. “Preliminary Rating ASA International Bank of Ghana. 2011. Bank of Ghana Annual Report 2011. Holding.” Wayne, Penn.: GIIRS. www.giirs.org Accra: Bank of Ghana. Banque Centrale Du Congo. 2012. “Aggregated Financial IMF (International Monetary Fund). 2011. Ghana: Financial Sector Data.” Kinshasa: Banque Centrale Du Congo, System Stability Assessment Update 2011. Washington, December. D.C.: IMF. http://www.imf.org/external/pubs/ft/ BCM (Banque Centrale de Madagascar). 2012a. “Regular scr/2011/cr11131.pdf Note on the Evolution of the Activity of Microfinance INSTAT. 2011. “The Labour Market in the Agglomoration Institutions at the End of December.” Antananarivo: of Antananarivo in 2010: A Decenium Perspective.” BCM. Antananarivo: INSTAT. ———. 2012b. “Bulletin of the Central Bank of Madagascar.” Jansson, T. 2010. “Success Factors in Microfinance Number 21. Antananarivo: BCM, June. Greenfielding.” IFC SmartLessons. Washington, D.C.: ———. 2012c. “Regular Note on the Evolution of the Activity IFC, May. of Banks and Financial Establishment at the End of McKee, Katharine, Estelle Lahaye, and Antonique Koning. December, 2012.” Antananarivo: BCM. 2011. “Responsible Finance: Putting Principles to Work.” Beck, T., S. M. Maimbo, I. Faye, and T. Triki. 2011. “Financing Focus Note 73. Washington, D.C.: CGAP. Africa: Through the Crisis and beyond.” MFW4A. MicroCred Madagascar. 2012. “Activity Report 2011.” Washington, D.C.: World Bank, BMZ, and AfDB. Antananarivo: MicroCred. Buchs, T. and J. Mathisen. “Competition and Efficiency in Moyart, M., et al. 2005. “Country-Level Effectiveness of Aid Banking: Behavioral Evidence from Ghana.” IMF Report (CLEAR) for Madagascar.” Washington, D.C.: Working Paper 2005. Washington, D.C.: IMF. CGAP. Cabinet ECR. 2011. “Study and Analysis of the Legal, Reglementary and Fiscal Framework of the Microfinance Nair, A., and A. Fissha. 2010. “Rural Banking—The Case of Sector in Madagascar.” Antananarivo: Cabinet ECR. Rural and Community Banks in Ghana.” Agriculture and Rural Development Discussion Paper 48. Washington, CNMF. 2007. “National Microfinance Strategy 2008–2012.” D.C.: World Bank, May. Antananarivo: CNMF. ———. 2012. “Consolidated Statistics of Microfinance Porteous, David. 2006. “Competition and Microcredit Activities, December 2012.” Preliminary version. Interest Rates.” Focus Note 33. Washington, D.C.: CGAP, Antananarivo: CNMF. February. 37 ProCredit Holding. 2010. Annual Report. Frankfurt am Schmidt, Reinhard H. The View of the Participating Main: ProCredit Holding. Observer: An Outline of the IPC/IMI Story from ———. 2012. Annual Report ProCredit Bank Congo 2011. Neoinstitutional Perspective, in Economy under Latin Frankfurt am Main: ProCredit Holding. American Conditions. Frankfurt am Main: Vervuert, pp. Rosenberg, Richard, Scott Gaul, William Ford, and Olga 95–122. Tomilova. 2013. “Microcredit Interest Rates and Their Transition Presidency. 2013. “Law n°2002-021 covering Determinants, 2004–2011.” Forum 7. Washington, D.C.: the laws for finance for 2013.” Government of CGAP, June. Madagascar. 38