94365 SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE i SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE Contents Acknowledgements 1 Acronyms 2 Executive Summary 3 1. INTRODUCTION 9 2. MSMEs IN ETHIOPIA: THE MACRO ECONOMIC CONTEXT AND INSTITUTIONAL FRAMEWORK 17 2.1 Strategic and macroeconomic context for MSME GROWTH 17 2.2 Legal and Regulatory Framework 18 2.3 Outlook of The Ethiopian Financial Sector 19 3. DEMAND-SIDE ANALYSIS 23 3.1. The place of small and medium enterprises in the economy 24 3.2 Is access to finance an obstacle to business? 28 3.2.1 Firms’ perceptions of access to finance in Ethiopia 28 3.2.2 More than elsewhere firms are considerably credit constrained 30 3.3 Liquidity Constraints 32 4. SUPPLY-SIDE ANALYSIS 35 4.1. Defining MSME Financing in Ethiopia 38 4.2. The Extent of Banks’ and MFIs’ involvement with SMES 39 4.3. Drivers and obstacles to SME Financing 43 4.3.1. Drivers of SME Finance 43 4.3.2. Obstacles to SME Financing 46 4.4. The Banks’ and MFIs’ Business Models 48 4.4.1. Organizational Models 48 4.4.2. MSME Specific Products and Marketing 50 5. INTERNATIONAL BEST PRACTICES IN SUPPORTING SME FINANCE 53 5.1. Key Findings on SME Finance Practices in Ethiopia vis-à-vis International Best Practices 53 5.1.1. The importance of a commonly agreed definition of MSMEs 53 5.1.2.The existence of a missing middle phenomenon 54 5.1.3. The excessive collateral requirements 55 5.1.4. The role of Government programs 56 5.1.5. The inadequateness of business models. 57 SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE v 5.2. Public Support Programs 57 5.2.1. Training Opportunities 58 5.2.2. Technology Transfer 58 5.2.3. Leasing Services 58 5.2.4. Facilities 58 5.2.5. Capacity Building 58 5.2.6. Partial Credit Guarantee Schemes 58 5.2.7. MFI Up-Scaling Intervention: The Women Entrepreneurship Development Project (WEDP) 59 5.3. Alternative Sources of Financing 60 5.3.1. Leasing 60 5.3.2. Factoring 60 5.3.3. Capital Markets 60 5.4. Other International Support Programs 61 6. POLICY CONSIDERATIONS AND POTENTIAL POLICY DIRECTIONS 63 Conclusions 67 References 69 Appendix 1: About the Enterprise Surveys 71 Appendix 2: About the Survey of Large and Medium Scale Manufacturing Industries (LMMIS) 73 Appendix 3: Defining Degrees of Credit Constraint Status 75 iquidity constraints Appendix 4: L 76 vi Figures Figure 1. Tackling the missing middle from both sides 5 Figure 2. Enabling environment for SME finance 6 Figure 3. Ethiopia: Half of Engaged Persons in the Industrial Sector are in Micro Establishments (2007/08) 9 Figure 4. SMEs* are the most constrained by credit in the Sub-Saharan Africa region 10 Figure 5. Ethiopia: Private sector credit (% of GDP) 20 Figure 6. Ethiopia: profitability of the banking sector (ROE) 21 Figure 7. Ethiopia: Access to Finance is a Top Obstacle 22 Figure 8. Employment Trends in Manufacturing (2000-2011), by Age and Size Groups 26 Figure 9. Net New Jobs (2009-2011), by Industry 27 Figure 10. Annual Employment Growth Rate (%) (2009-2011), by Industry 28 Figure 11. Access to Finance is a larger obstacle for young and small firms 29 Figure 12: Nearly half of firms in Ethiopia are fully credit constrained 31 Figure 13. Most Promising Segments for Growth 39 Figure 14. MFIs Gross Deposits and Outstanding Loans in Billion Birr (Dec. 31, 2012) 40 Figure 15. Banks Gross Deposits and Outstanding Loans in Billion Birr (Dec. 31, 2012) 40 Figure 16. Share of SMEs Lending in Overall Portfolio 41 Figure 17. The Missing Middle: Lending to SMEs is limited 41 Figure 18: Proportion of Outstanding Deposits, by Client Size 42 Figure 19. Average Interest Rates as at December 2012 43 Figure 20. The Main Drivers of banks’ and MFIs’ involvement with SMEs and Micro Enterprises 44 Figure 21. Banks & MFIs: Comparison of Cost, Risk and Profitability of SME Loans versus Large Enterprise Loans 44 Figure 22. Burden posed by regulatory documentation requirements for lending to MSMEs 45 Figure 23. Impact of government programs on SME finance 45 Figure 24. Importance of credit bureaus for Banks’ MSME finance 46 Figure 25. Obstacles to Banks & MFIs Involvement with SMEs (Significant and Very Significant) 47 Figure 26. Having a separate MSME department/unit to manage SMEs and Micro Enterprise clients 48 Figure 27. Organization of credit risk function (Banks and MFIs) 48 Figure 28. Loan origination and monitoring of micro-enterprise and SME loans 49 Figure 29. Use of qualitative assessment and variables for credit analysis of micro-enterprise and SME loans 50 Figure 30. Use of quantitative assessment and variables for credit analysis of micro-enterprise and SME loans 50 Figure 31. Marketing Focus of MFIs and Banks 52 Figure 32. Banks & MFIs distribution channels 52 Figure 33. Tackling the missing middle phenomenon 55 Figure 34. Value of collateral needed for a loan (% of the loan amount). 55 Figure 35. MSME finance technology and characteristics 56 Figure 36. Public Support Schemes and Private Sector Initiatives by Sub-Categories 56 Figure 37. MSME Banking Value Chain 57 Figure 38. Flow of funds in the WEDP line of credit 59 SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE vii Figure 39. The SME Finance Productivity Frontier 64 Figure 40. ES Sample Distribution, by Firm Size and Region 71 Figure 41. ES Sample Distribution, by Industry 71 Table 22: Investment on cash flow – Small manufacturing firms from 2000 to 2011 81 Tables Table 1. Sectoral Contribution to GDP and GDP Growth (in Million Birr) 18 Table 2. Loans and Advances by Lenders (in Million Birr) 20 Table 3. Firm Size Transition in Manufacturing (t=2008-2010) 25 Table 4. Reasons for Not Applying For a Loan (% of Firms by Size Group) 30 Table 5. Types of Collateral Used 30 Table 6. Performance and Credit Constraint 32 Table 7. MSME Definitions, by National MSE Development Strategy 38 Table 8. MFIs definition of Micro and SMEs, by turnover, employee size, and loan size 39 Table 9. Total Lending (Dec. 2012, Birr) 42 Table 10. Number of Loans Outstanding (Dec. 31, 2012) 42 Table 11. Business models of banks and MFIs 49 Table 12. Product and Services Offered, by MFIs and Banks 51 Table 13. Target markets of banks and MFIs 51 Table 14. Selected MSME definitions across the world 54 Table 15. Enhancing Finance in China and Turkey- Project Indicators and Financial Conditions 61 Table 16. LMMIS Sample Size 73 Table 18: Investment on cash flow – Manufacturing firms from 2000 to 2011: The choice of the Young variable 77 Table 20: Number of observations over 2000-2011 79 Table 21: Investment on cash flow – Manufacturing firms from 2000 to 2011 80 Table 23: Investment on cash flow – Medium manufacturing firms from 2000 to 2011 82 Table 24: Investment on cash flow – Large manufacturing firms from 2000 to 2011 83 viii Acknowledgements The project team would like to thank the National Authorities of the Federal Democratic Republic of Ethiopia and all participants in the surveys and in the consultation workshop held in February 2014 who provided valuable inputs related to the study. In particular, we would like to thank representatives from Ministry of Finance and Economic Development, National Bank of Ethiopia, Development Bank of Ethiopia, Federal Micro and Small Enterprises Development Agency, Commercial Bank of Ethiopia, Dashen Bank, NIB Bank, Awash International Bank, Construction and Business Bank, Wegagen Bank, Amhara Credit and Saving Institution (ACSI), Oromia Credit and Saving Share Company (OCSSCO), Addis Credit and Saving Institution (AdSCI), Omo Micro Finance Institution, Wasasa, Wisdom MFI and surveyed micro, small and medium enterprises. Additionally, we would like to thank Irina Astrakhan, Lars Moller, Asya Akhlaque, Senidu Fanuel, Mehnaz Safavian, Silvia Muzi, Andrej Popovic, Xavier Cirera, Bilal Zia for helpful comments and suggestions on earlier drafts of this report. The study was carried out under the overall guidance of Gaiv Tata (Sector Director, Financial and Private Sector Development, Africa Region, World Bank) and Guang Zhe Chen (Country Director for Ethiopia, World Bank). This report was led by Francesco Strobbe (Senior Financial Economist, World Bank). Team members included Johanne Buba (JPO, World Bank), Leonardo Iacovone (Senior Economist, World Bank), Mathewos Shamo (Consultant, World Bank), Lisa Stahl (JPO, World Bank), Eyob Tolina (Consultant, World Bank) and Judy Yang (Consultant, World Bank). Naturally all errors remain our own. SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 1 Acronyms CBE Commercial Bank of Ethiopia CSA Central Statistical Agency DBE Development Bank of Ethiopia ES Enterprise Surveys LIC Low-Income Countries GTP Growth and Transformation Plan ICPAE The Institute of Certified Public Accountants of Ethiopia IFAC The International Federation of Accountants LMMIS Large and Medium Scale Manufacturing Industries MDG Millennium Development Goals MFI Microfinance institutions MoFED Ministry of Finance and Economic Development MSME Micro, Small and Medium Enterprises NAABE National Accounting and Auditing Board in Ethiopia NBE National Bank of Ethiopia PASDEP Plan for Accelerated and Sustainable Development to End Poverty PCG Partial Credit Guarantee SEZ Special Economic Zone SME Small and Medium Enterprises SSA Sub-Saharan Africa Region 2 Executive Summary The Growth Transformation Plan (GTP, 2010/11-2014/15) prepared by the Government of Ethiopia (GoE) is the medium term strategic framework for a five-year period that shall guide the country’s efforts towards accelerating GDP growth, employment creation and transforming Ethiopia from a predominantly agrarian to a modern and industrialized economy. The GTP aims to increase the per capita income of its citizens to middle-income levels by 2025. This is just one goal in a portfolio of important socioeconomic targets for the future. Private sector development strategies to reach these goals include enhancing the productivity and modernization of Ethiopia’s agricultural sector, and enhancing the technological sophistication and economic contribution of the industrial sector. Ambitious targets have been set for the industrial sector and the GTP specifically identifies micro and small enterprises (MSEs) development as the key industrial policy direction for creating employment opportunities in Ethiopia.1 In particular, MSEs are expected to contribute greatly towards the GTP targets through the creation of more than 3 million jobs over the 5-year reference period.2 But are MSEs currently the main net job creators in Ethiopia? Are they able to access adequate financial services in a context where the banking sector is registering profitability levels well above the Sub-Saharan African average?3 These are some of the key questions that the present study aims at addressing. At the international level, the key role played especially by small and medium sized enterprises in economic development and their contribution to economic diversification and employment is widely recognized, and so is the reality that these enterprises face financing constraints around the world, both in developed and developing markets (Ayyagari et al. 2007; Beck et al. 2005). This study starts with a brief analysis of which firms are the main net job creators in Ethiopia and then focuses on the financing constraints of Ethiopian MSMEs as one of the key obstacles to job creation and growth.4 In doing so, the study uses two demand-side surveys (the Ethiopia Survey of Large and Medium Scale Manufacturing Industries – LMMIS, an unbalanced panel composed of about 6,000 firms with at least 10 employees which allows for a study of firm dynamics from 2000 through 2011 and the World Bank’s Enterprise Survey (ES) that was conducted between July 2011 and July 2012 and includes 794 firms which allows for the additional examination of the services sector, microenterprises, and a more detailed understanding of firm experiences with respect to access to finance) and an ad-hoc supply side survey administered to 16 financial institutions, including the major public and private sector commercial banks and microfinance institutions, covering over 90% of the total assets in the banking and microfinance sector.5 This survey allowed collecting data on the actual involvement of financial institutions with MSMEs, their perception of potential public policy approaches to enhance MSME access to finance and the adequateness of their current business models. The combination of both demand-side and supply-side analysis allows to gain a full picture of MSME finance practices in Ethiopia by connecting information on firm experiences with the reporting of financial institutions on their business practices. While there was already anecdotal evidence that small firms were lacking proper access to finance in Ethiopia, the value added of this study is to provide accurate empirical evidence of the existence of a missing middle phenomenon. 1 The industrial sector has been identified as the leading sector for realizing the country’s development objectives. Industry’s share in overall GDP is expected to increase from 12.9% in 2009/10 to 18.8% by 2014/15. 2 Under the leadership of FeMSEDA, support to MSEs has been extensive, and has helped to achieve targets set out in the GTP. From 2010 to 2013, 578,005 MSEs were supported by FeMSEDA and received at total of 4,807,314,875 Birr in financing ($240 Million USD), creating an estimated 1,024,739 jobs. (Source: FeMSEDA MSE Yearly Statistical Bulletin, 2013). 3 See Figure 4 on page 29. 4 While there is no universally agreed definition of MSMEs, the supply side analysis of this study makes use of the definition for MSEs contained in the 2011 Ethiopia National MSE Strategy (i.e. by focusing on head count only, microenterprises have up to 5 employees, small enterprises have from 6 to 30 employees) while the demand-side analysis relies on the definition used by the Ethiopian CSA’s Survey of Large and Medium Scale Manufacturing Industries (i.e. micro as 0-9 employees, small as 10-20 employees, medium as 21-99 employees and large as 100+ employees). 5 Responses to the supply-side survey were provided by thirteen out of sixteen financial institutions: seven banks representing 87.1 percent of the banking sector asset portfolio and six microfinance institutions representing 70 percent of the micro finance sector asset portfolio. SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 3 Findings from both demand-side and supply-side surveys clearly indicate the existence of a missing middle phenomenon in Ethiopia whereby small enterprises are more credit constrained than either micro or medium/large enterprises. In particular, the demand-side analysis shows that MSMEs in Ethiopia perform much worse than large firms across a host of finance indicators. MSMEs are much more likely to be rejected for loans, and less likely to have a loan, line of credit, or overdraft facility. These firms are also more likely to avoid loan applications all together due to high collateral requirements. However, the data reveals that the lack of access to finance is even more severe for small firms than for microenterprises. Main findings from the demand-side can be summarized into the following three categories: i. Job creation and employment growth: Job creation and employment growth is concentrated in large established (i.e. older) firms in both service and manufacturing sectors. Moreover, job creation and employment in the services and retail sectors are higher than in the manufacturing sector. ii. Access to finance: overall, data indicates the existence of a missing middle phenomenon in terms of financial services catering to small firms. Young and smaller firms are much more likely to be rejected for a loan or a line of credit than firms who are more established or larger. Moreover, despite confirming their need for improved access to finance, SMEs are discouraged from applying for loans due to excessively high collateral requirements. iii. Firms’ performance: Firms that are credit constrained exhibit poorer performance and productivity. The supply-side analysis confirms that small and medium enterprises are being underserved compared to micro and large firms. Micro finance institutions (MFIs) primarily cater to micro firms, leaving small and medium firms financially excluded. Large banks are discouraged from serving this segment primarily because of perceptions of lower returns and higher risk. However, most MFIs and banks also view the SME segment as the most promising one for growth and having good prospects. Key findings from the supply-side survey include the following6: i. SME finance culture: Financial institutions lack an “SME finance culture”: a harmonized definition of MSMEs is missing and consequently specific MSME financing strategies are not in place. ii. Perception of SMEs’ market segment: Financial institutions believe that market potential is very good for the SME segment. Expected returns and the contribution to the economic development of the country are seen as main drivers for lending to SMEs and microenterprises. iii. Risks and obstacles to SME finance: SME-specific factors and macroeconomic factors were indicated by all MFIs and banks as significant obstacles to the development of SME lending. While banks and MFIs believe there is high growth potential in lending for small enterprises, the current lack of involvement is due to perceived risks. iv. The missing middle phenomenon: SMEs represents a missing middle in the financial sector with high heterogeneity of lending patterns between MFIs and commercial banks. v. Business models: The business models of the surveyed financial institutions are mostly inadequate to serve SMEs: they lack a dedicated and specialized SME unit or department within their organizational structure; loan appraisal techniques are still mostly based on traditional relationship lending rather than on transactional technologies such as credit scoring etc.; products are highly standardized and there is very limited product innovation; distribution channels are still largely based on branches and long term financing needs of SMEs do not seem to be properly addressed. 6 Details on the specific key findings from the demand-side and supply-side analysis are summarized in the Introduction and further developed in the reminder of the report. 4 vi. The importance of Government financed programs: Government financed programs (viii) Regulatory and judicial issues. The contractual environment and lack of collateral (credit guarantee programs and line of credit with technical assistance) are important registry inhibit secured lending and constrain access to finance for SMEs. drivers. Changes in the market due to imposed lending restrictions are seen as a hindrance causing liquidity constraints. vii. A supportive financial sector infrastructure. The potentialities of the credit bureau are not effectively exploited. Based on these analytical findings, the report proposes a series of potential policy directions and recommendations viii. on how Regulatory and judicial to The issues. improve access contractual to financeand environment forlack SMEs, particularly of collateral registry to ensure small enterprises,inhibit their secured contribution lending to Ethiopia’s and constrain Growth access to finance and for Transformation Changes in the Plan. SMEs. market Seven specific policy directions have been identified and are summarized in the table below. due to imposed lending restrictions are seen as a hindrance causing liquidity constraints. However, for easiness of interpretation, those seven policy directions could be further Based on these analytical findings, the report proposes a series of potential policy directions and grouped into two main categories: recommendations on how to improve access to finance for SMEs, particularly small enterprises, to ensure their contribution to Ethiopia’s Growth and Transformation Plan. Seven specific policy i. Categoryhave directions set of 1: abeen interventions identified to help commercial and are summarized in the table banks to downscale below. However, theirof for easiness business (i.e. through the joint provision of dedicated liquidity plus tailored technical interpretation, those seven policy directions could be further grouped into two main categories: assistance to participating banks on the supply side and the provision of business Category development1: a set of interventions skills to helpSMEs training to address commercial banks specific to downscale constraints on the their business demand (i.e. side). through the joint provision of dedicated liquidity plus tailored technical This intervention would complement the current effort that the GoE is alreadyassistance to participating banks on the supply undertaking by side and the provision supporting MFIs to upscale development of business skills training their business under totheaddress Women SMEs specific constraints on the demand side). This intervention would complement the current effort Entrepreneurship Development Project (WEDP). Figure 1 shows how the combined that the GoE is already undertaking by supporting MFIs to upscale their business under the Women action of MFIs upscaling and commercial banks’ downscaling can simultaneously Entrepreneurship Development Project (WEDP). Figure 1 shows how the combined action of MFIs tackle the upscaling andmissing middle commercial with banks’ MFIs serving downscaling the lower bound can simultaneously of the tackle small enterprise themissing middle with segment and commercial banks serving the upper bound. MFIs serving the lower bound of the small enterprise segment and commercial banks serving the upper bound. Figure 1. Tackling the missing middle from both sides Figure 1. Tackling the missing middle from both sides How?     Dedicated   ng ks   cali liquidity  (line   n s Medium   Ba own of  credit) Enterprises   D + TA  to  banks + Missing  Middle:  Small   BDS  to  SMEs Enterprises   M Up FIs   Micro  Enterprises   sc a ling WEDP  (currently   ongoing) Category 2: a set of interventions to further improve the enabling environment for facilitating SME finance, particularly with a focus on the insolvency and creditor/debtor regime on one side and on the market credit information on the other side, as shown in Figure 2 below. 12 SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 5 ii. Category 2: a set of interventions to further improve the enabling environment for facilitating SME finance, particularly with a focus on the insolvency and creditor/debtor regime on one side and on the market credit information on the other side, as shown in Figure 2 below. Figure 2. Enabling environment Figure 2. Enabling for environment SME for finance SME finance Credit Access / Credit Risk Protection   Management   Insolvency   Collateral Systems Credit Liquidation Information immovable / systems Reorganization movable Enforcement Systems Risk Insolvency of Management Enterprise Public Auction & Practices   groups   Collections   The following table provides additional details on the specific seven potential policy directions identified: The following table provides additional details on the specific seven potential policy directions identified: Potential policy direction Description 1. Foster an SME finance culture QQ A commonly agreed and harmonized definition of what constitutes an Potential policy financial institutions and Description direction among SME is crucial for preparing any type of support intervention or strategy. stakeholders 1. Foster an SME finance culture • A commonly QQ The agreed and current definition forharmonized definition micro and small of what enterprises constitutes contained in the among financial institutions and an SME isMSE “National crucial for preparing Development any type Strategy of support (2011)” intervention starting or is a good point, strategy. shared by most MFIs. However setting up a stakeholders’ working group stakeholders The current definition for microaand small enterprises contained • to look into the issue of finding common definition to be usedin by all the “National MSE Development Strategy (2011)” is a good financial institutions to segment the market would be recommended. starting point, shared by most MFIs. However setting up a stakeholders’ QQ In working turn, a commonly to look into groupdefinition agreed would thebe issue of finding beneficial a for regulatory common definition to be used by all financial institutions to and reporting purposes allowing for proper analysis of SME finance segment the market would be recommended. development over time. • In turn, a commonly agreed definition would be beneficial for 2. Complement current QQ The positive role that public support interventions can play in promoting regulatory and reporting purposes allowing for proper analysis of Government support programs SMEfinance SME finance practices came development clearly from the analysis. outtime. over 2. Complement current for MSMEs withGovernment the promotion • The positive role that public support interventions can play in QQ While the Government of Ethiopia is currently engaged in a series of of commercial support programs for MSMEs banks’with promoting SME finance practices came out clearly from the downscaling interventions analysis. efforts to promote the engagement of MFIs in the small business positive the promotion of commercial While thethrough • segment Governmentmarket Ethiopia isefforts of up-scaling (e.g. currently WEDP),in engaged the banking a series banks’ downscaling interventions sector requires of positive further efforts thein attention to promote order to exploit engagement of MFIsits potential in serving in the small business the missingsegment middle through segment.market up-scaling efforts (e.g. WEDP), the banking sector requires further attention in order to exploit its QQ Incentives should be provided to commercial banks for engaging in market downscaling initiatives. Successful examples of international best 13 practices showed that the combination of dedicated liquidity (through lines of credit) with tailored technical assistance packages prove to be effective in successfully reaching the missing middle segment. QQ In Ethiopia, institutions like the Development Bank of Ethiopia are currently well placed to play the wholesaler role in providing dedicated liquidity for SME finance to financial intermediaries. 6 Potential policy direction Description 3. Promote innovation in QQ Around the globe, certain commercial banks have applied lending financial products and lending practices developed in the microfinance sector to overcome the issues of technologies by providing high transaction costs and high-risk profiles of potential borrowers. These incentives to commercial banks best practices can be of use in the Ethiopian context as well. through tailored technical assistance. QQ The surveys revealed that most Ethiopian financial institutions do not have the right instruments to put in place an innovative product mix and to engage in new lending technologies. QQ Tailored technical assistance, coupled with the provision of dedicated liquidity as under point #2 above, would help to stimulate the use of new techniques in line with international best practices. It would help commercial banks to establish dedicated SME units within their organization, define SME-specific strategies, offer a range of products beyond lending, utilize low cost delivery channels, develop and use risk modeling tools and build adequate hardware and software architecture. 4. Address SME specific factors QQ BDS can help to address some of the intrinsic weaknesses in SMEs that through the provision cannot be addressed through financing tools alone. of adequate business development skills in QQ The survey indicates that SME specific weaknesses are among the major conjunction with interventions factors inhibiting commercial banks to engage in this market segment. on the supply-side. QQ The government has a role in this market as a provider of BDS as it is already currently doing through FeMSEDA and the TVETs. However, it would be crucial to coordinate and tailor BDS interventions with parallel incentives provided to the supply side (i.e. commercial banks) to engage in SME lending. 5. Promote policies aimed at QQ Given the prevalence of collateral lending in Ethiopia, establishing a addressing the limitations collateral registry of both movable and immovable assets is important of the current collateral for creating an effective credit market by expanding the scope of secured regime based on an accurate lending transactions and improving access to financial services. The lack of diagnostic of the Insolvency a collateral registry in fact is currently a major obstacle to the promotion and Creditor/Debtor of SME finance. regimes. QQ Policies to facilitate the setting-up of the movable assets registry would certainly be beneficial as currently collateral rates in Ethiopia are among the highest in the region. The proposed registry would document charges and collaterals created by borrowers to secure credit facilities provided by lenders. QQ The legislation related to the contractual environment would also benefit from a thorough analysis. Particularly it would be recommended to conduct a diagnostic of the creditor rights and enforcement systems (for secured and unsecured credit); credit risk management, debt recovery and informal enterprise workout practices; formal insolvency system (liquidation and reorganization proceedings); and effectiveness of the relevant institutional and regulatory frameworks in implementing laws in this area. A similar diagnostic would allow to identify bottlenecks, facilitate access to credit for SMEs and provide a stable backdrop for private transactions. SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 7 Potential policy direction Description 6. Support development of QQ While comments generally were positive about information from market credit information for the credit bureau, it was noted that coverage is limited and does not SMEs necessarily include all financing received by SMEs. QQ Currently, Ethiopia has a credit information system with high technological capabilities and business functionalities to run credit reporting. However, it is being used merely as a database management system by the National Bank of Ethiopia that collects information on creditworthiness of borrowers from supervised financial institutions, makes such information available for financial institutions, and uses it primarily for supervisory purposes. QQ Enhancing its functions to provide reliable and value added credit information, both positive and negative information (full file), is critically important for financial institutions for quality decision making on credit, risk mitigation and minimizing fraud by providing value added services such as credit scoring, marketing service, application processing and portfolio monitoring. QQ Specific policies could support the use of market credit information through the refinements to the legal and regulatory framework to improve incentives to share information among lenders, and an educational campaign promoting the value of credit bureaus among SMEs and financial institutions. 7. Develop the institutional QQ Among the alternative sources of funding, the Leasing Proclamation has framework for alternative been amended recently by providing licensing and supervisory authority sources of funding. to the National Bank of Ethiopia to finance leasing and hire-purchase leasing. Leasing can be an effective mechanism for boosting the Ethiopian economy by providing long-term finance to SMEs and its use should be more extensively promoted. QQ Similarly, the institutional framework for other sources of funding such as factoring and joint venture capital need to be developed as a part of a comprehensive package of financial sector infrastructure and products development 8 1. INTRODUCTION The private sector plays a critical role as a catalyst for economic change through 1. INTRODUCTION offering financing options which allow firms to expand and innovate. In Ethiopia, small firms face more challenges in obtaining formal financing than large firms; they are much more likely to be rejected for loans, and are less likely to have external financing. Banks The private sector primarily cater toplays largea critical role as firms and, a catalyst although for economic they perceive thechange SMEthrough segment offering financing as a promising options which allow firms to expand and innovate. In Ethiopia, small firms one in terms of growth prospects, they also tend to see SME lending as having higher risks face more challenges obtaining inand formal financing lower profitability than large than lending they are much more likely to be rejected for loans, firms; enterprises. to large and are less likely to have external financing. Banks primarily cater to large firms and, although The role they perceive of micro the SME segment(0-9asemployees) a promising andonesmall (10-20 in terms employees) of growth enterprises prospects, will they also be tend to see SME crucial tolending as having reach growth higher targets theand risks for lower profitability industrial than lending sector as outlined to large enterprises. by Ethiopia’s five-year Growth and Transformation Plan (GTP) (2010/11-2014/15). GTP’s policy in Trade and The role of includes Industry micro (0-9 employees) providing and small special attention(10-20 employees) to micro enterprises and small will be enterprises crucialtraining to reach through growth targets for the industrial sector as outlined by Ethiopia’s five-year Growth and promoting increased hiring. Over the five-year period, 3.4 million jobs are expected to be and Transformation Plan to(2010/11-2014/15). (GTP) added this segment, and GTP’s trainingpolicy in Trade to over and Industry 3 million includes operators providing to be special conducted. attention In 2007/08, to micro and small enterprises through training and promoting increased hiring. 96 percent of all firms in Ethiopia’s industrial sector were micro, and over half of all engaged Over the five-year period, 3.4 million jobs are expected to be added persons were affiliated with microenterprises (Figure 3). to this segment, and training to over 3 million operators to be conducted. In 2007/08, 96 percent of all firms in Ethiopia’s industrial sector were micro, and over half of all engaged persons were affiliated with microenterprises (Figure 3). Figure Figure 3. Ethiopia: 3. Ethiopia: Half of Half of Engaged Engaged Persons Persons in the in the Industrial Sector are in7Micro Industrial 7 Sector are in Micro Establishments (2007/08) Establishments (2007/08) Micro (0-9 employees) SMEs & Large (>10 employees) 100% 1,926 132,172 130,305 0% Number of Establishments Number of Persons Engaged Number of Employees Notes: Number Engaged: includes paid employees, unpaid working proprietors, active partners, unpaid family workers Notes: and paid Number and unpaid Engaged: apprentices. includes paidMicro (ie. Small employees, Scale) unpaid Manufacturing working Establishments: proprietors, engaging active partners, less unpaid than 10 persons family and use workers power and paid-driven machinery. and unpaid apprentices. Micro (ie. Small Scale) Manufacturing Establishments: Source: Report on engaging Large less and than 10Medium persons Scale Manufacturing and use and Electricity power -driven machinery. Industries Survey 2011, Report on Small Scale Manufacturing Industries Survey 2010 Source: Report on Large and Medium Scale Manufacturing and Electricity Industries Survey 2011, Report on Small Scale Manufacturing Industries Survey 2010 However, an analysis we conducted on a panel of 6000 Ethiopian firms from 2000 to 2011, reveals that large and more established (i.e. older) firms are more important as net job creators than small and younger firms. 7 Micro-establishments are also referred to as small-scale firms by the CSA. Therefore, the present study, while keeping job creation as its starting point, has looked into 16 financing constraints of SMEs as one of the possible main reasons preventing SMEs in Ethiopia from playing their catalyst role. SME financing, in fact, is now at the centre of the international development agenda and is of considerable interest to policy makers due to their importance for economic development as well as their potential contribution to economic diversification and job creation. However, SME growth potential in developing economies, especially in Africa, is limited as they are significantly more credit constrained compared to larger enterprises8. Kuntchev, Ramalho, Rodriguez-Meza, and Yang (2012) find that over half of SMEs are credit constrained in the Sub-Saharan Africa (SSA) region, which is higher than in any other region9 (Figure 4). 7 Micro-establishments are also referred to as small-scale firms by the CSA. 8 Fuchs and Berg 2013; Beck and Demigurc-Kunt 2006; Beck et al 2006; Beck, Demirgüç-Kunt & Maksimovic 2005, and Beck, Demirgüç-Kunt, Laeven, Maksimovic 2006 9 The Middle East and North Africa region is excluded since only Yemen had been surveyed in that region at the time of the report. SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 9 development as well as their potential contribution to economic diversification and job creation. However, SME growth potential in developing economies, especially in Africa, is limited as they are significantly more credit constrained compared to larger enterprises8. Kuntchev, Ramalho, Rodriguez-Meza, and Yang (2012) find that over half of SMEs are credit constrained in the Sub-Saharan Africa (SSA) region, which is higher than in any other region9 (Figure 4). The proportion of SMEs in Ethiopia that are credit constrained is even slightly higher than the SSA average. Access to finance is important since the use of external finance among The SMEsproportion of SMEs is associated in Ethiopia with greater that are credit innovation constrained (GFDR, is even 2014). slightly higher In Sub-Saharan than the Africa, SSA banks average. Access to finance is important since the use of external finance among SMEs have a crucial role to play in addressing this credit constraint obstacle due to their dominance is associated with greater in the innovation financial systems(GFDR, and 2014). the lackIn Sub-Saharan Africa, banks of SME financing providedhaveby a crucial role financing to play in informal addressing mechanisms. this credit constraint obstacle due to their dominance in the financial systems and the lack of SME financing provided by informal financing mechanisms. Figure 4. SMEs* are the most constrained by credit in the Sub-Saharan Africa region Figure 4. SMEs* are the most constrained by credit in the Sub-Saharan Africa region 60 50 % of SMEs that are Credit 40 Constrained 30 56 55 20 38 39 29 33 10 0 Ethiopia SSA EAP ECA LAC SAR *SMEs defined by the Enterprise Surveys are establishments between 5-99 employees. Enterprise Surveys does not include microenterprises *SMEs defined by the (0-4 employees). Enterprise Surveys are establishments between 5-99 employees. Enterprise Notes: Countries are grouped Surveys does per region not include according to(0-4 microenterprises the World Bank classification. employees). Source: Kuntchev, Ramalho, Rodriguez-Meza, and Yang (2012), Enterprise Surveys Notes: Countries are grouped per region according to the World Bank classification. Source: Kuntchev, 8 Ramalho, Fuchs and Berg 2013; BeckRodriguez-Meza, and and Demigurc-Kunt Yang Beck 2006; (2012), et Enterprise Surveys al 2006; Beck, Demirgüç-Kunt & Maksimovic 2005, and Beck, Demirgüç-Kunt, Laeven, Maksimovic 2006 9 analyze financing conditions for MSMEs in Ethiopia, this study utilizes unique information from To The Middle East and North Africa region is excluded since only Yemen had been surveyed in that region at three surveys. the time The Ethiopia’s Survey of Large and Medium Scale Manufacturing Industries (2000- of the report. 2011) and the World Bank’s Enterprise Survey that was conducted in Ethiopia in 2011 are used for the demand-side analysis. Specifically the former 17dataset provides insights on firm dynamics while the latter helps to capture the experiences of firms in the services and retail sectors and to have a more detailed view of firm financing experiences. Moreover, in order to capture the perspectives and drivers of financing institutions, an ad-hoc supply-side survey was issued to 16 financial institutions including the major public and private sector commercial banks and microfinance institutions covering over 90% of the total assets in the banking and microfinance sector. The survey allowed collecting data on the actual involvement of financial institutions with MSMEs, their perception of potential public policy approaches to enhance SME access to finance and the adequateness of their current business models. Studies on SME financing with demand and supply-side components have also been carried out in other Sub-Saharan Africa’s countries (including Kenya, Nigeria, Rwanda, South Africa, and Tanzania) in the North Africa and Middle East region and in the Latin America and Caribbean region (De la Torre et al. 2008; Beck et al. 2008; Fuchs et al, 2011; Rocha et al. 2010)10. While using a similar instrument to analyse the supply-side, this study differs from the previous ones conducted in Sub-Saharan Africa for a more comprehensive and detailed analysis on the demand-side which provides useful complementary recommendations to the more traditional ones based only on supply-side constraints. The combination of information on firm experiences with the reporting of financial institutions on their business practices, allowed us to gain a full picture of SME finance practices and perspectives in Ethiopia. 10 The financial crisis in 2008 also motivated the study to be conducted in several countries in the Latin America and Caribbean region evaluating the impact of SME finance programs (Acevedo/Tan, 2010). 10 The results from both demand-side and supply-side analysis clearly indicate the existence of a missing-middle phenomenon in Ethiopia. This is a common feature to many developing countries that have a large number of microenterprises and some large firms, but far fewer small and medium enterprises. In high-income countries, SMEs are responsible for over 50% of GDP and over 60% of employment, but in low-income countries they are less than half of that: 30% of employment and 17% of GDP.11 This SME gap is called the ‘missing middle’. Evidence from international research clearly shows that returns to capital are high in this segment.12 SMEs aren’t missing because they would not be profitable: they are missing because finance is not reaching them in an effective way. This shows that access to finance is a significant barrier, and that there is a massive profit opportunity for those who are able to successfully finance these firms. The SME finance studies conducted in other African countries are useful for benchmarking Ethiopia’s position with regards to access to finance. The share of SME lending in the overall portfolio of banks in the 5 countries where a similar study on the supply-side was conducted varies between 5 and 20 percent. Banks in Kenya, Rwanda, and Tanzania are more involved in SME lending in terms of the share of their portfolio than banks in South Africa and Nigeria. SME lending in Ethiopia is on the lower side, with SME lending comprising only 7 percent of bank portfolios. The studies found that major contributing factors for SME lending are the structure and size of the economy, the extent of government borrowing, the state of financial sector infrastructure and the enabling environment including government support programs. In the case of Ethiopia, main findings from the demand-side can be summarized into the following three categories: i. Job creation and employment growth: Job creation and employment growth is concentrated in large established (i.e. older) firms in both service and manufacturing sectors. Moreover, job creation and employment in the services and retail sectors are higher than in the manufacturing sector. a. Jobs analysis in the manufacturing sector shows that the majority of paid employment is found in large enterprises. Interestingly, in the services sector the job dynamics are similar to the ones in the manufacturing sector as large established firms tend to be the most important net jobs creators during the period 2009-2011. b. Typically, young firms are a great source of job creation, but this trend is not seen in Ethiopia. Conditional on survival, young Ethiopian firms enjoy higher employment growth rates than established firms, but they also generate a smaller number of new jobs than older firms. In the rest of the world, young firms are principal job creators despite high exit rates (Haltinwanger et al 2011; Rijkers et al, 2013). The fact that older firms dominate net job creation in Ethiopia is worrisome since it suggests there is a lack of competitiveness and innovation in the private sector. c. The manufacturing sector plays a limited role in the overall Ethiopian economy, comprising only 4.2 percent of GDP in 2012/13. Therefore, it’s important to also understand employment trends in the services sector. The service sector created nine times more jobs than the manufacturing sector during the period 2009-2011. ii. Access to finance: overall, data indicates the existence of a missing middle phenomenon in terms of financial services catering to small firms. Young and smaller firms are much more likely to be rejected for a loan or a line of credit than firms who are more established or larger. Moreover, despite confirming their need for improved access to finance, SMEs are discouraged from applying for loans due to excessively high collateral requirements. a. Only 1.9 percent of small firms have a loan or line of credit. This rate is lower than among micro, medium, and large firms (6.0, 20.5, and 35.5 percent respectively). Fifty-seven percent of small firms are fully credit constrained, a rate higher than in any other size group. These statistics corroborate with assertions that small firms 11 Ayyagari, Beck, & Demirguc-Kunt. “Small- and medium-enterprises across the globe: a new database” 12 E.g Banerjee & Duflo (India), McKenzie & Woodruff (Mexico) SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 11 struggle the most in obtaining access to finance since MFIs cater to micro-sized firms, and commercial bank clientele are predominantly medium and large firms. b. Among firms who applied for a loan or line of credit in fiscal year 2011, 57.3 and 87.9 percent of applications submitted by micro and small firms respectively were rejected; this is in sharp contrast to the much lower 6.2 and 10.4 percent rejection rate experienced by medium and large firms. Fifty-six percent of young firms had their loan applications rejected compared to 33 percent of established firms. c. Results show that SMEs are less likely to say that they don’t need a loan, thus confirming the need for finance. However, SMEs are discouraged (or voluntarily exclude themselves) from applying for loans or lines of credit. About a third of small and medium firms reported that they did not apply for a loan or line of credit because collateral requirements were too high. Collateral rates in Ethiopia are much higher than in more developed African economies. For example, collateral rates are only 120.8 percent of the loan value in Kenya (2007) compared to 234 percent in Ethiopia. Small firms are also the most likely to use personal assets as a type of collateral (36.8 percent of small firms use this type). iii. Firms’ performance: Firms that are credit constrained exhibit poorer performance and productivity. a. In Ethiopia, a credit constrained firm has sales growth 15 percentage points lower, employment growth 5 percentage points lower, and labor productivity growth 11 percentage points lower than firms who are not credit constrained. b. Investment decisions of manufacturing firms in Ethiopia are heavily dependent on cash flows. To identify the existence of credit constraints we evaluate the extent to which firms’ investments are reliant on cash flow. We find some evidence that liquidity constraints are stronger for young small firms when compared to established small firms. On the other hand, there is limited evidence that young medium and large firms are more credit constrained than other medium and large firms. On the other hand, key findings from the supply-side survey include the following: i. SME finance culture: Financial institutions lack an “SME finance culture”: a harmonized definition of MSMEs is missing and consequently specific MSME financing strategies are not in place. a. Financial institutions in Ethiopia lack a commonly agreed definition of MSMEs. While the majority of MFIs uniformly use the MSE definition that is laid out in the Government’s National MSE Development Strategy, commercial banks do not seem to uniformly distinguish among small, medium, and large enterprises. b. The lack of a harmonized definition leads to a lack of market segmentation and ultimately a lack of in depth customer knowledge and proper business strategy to address this untapped market segment. ii. Perception of SMEs’ market segment: Financial institutions believe that market potential is very good for the SME segment. Expected returns and the contribution to the economic development of the country are seen as main drivers for lending to SMEs and microenterprises. a. The co-existence of these two motors, on the one hand the economic dimension of business profitability, and on the other hand the more political dimension of contribution to the country’s economic development, represents an interesting feature of the Ethiopian market. Namely, publicly owned financial institutions dominate both the banking and the microfinance sector. 12 b. Although MFI and bank involvement in SME lending is limited, these same financial institutions believe that the potential for this segment of the market is very good. The majority of surveyed financial institutions believe that prospects for the SME market are good and that the SME market size is large. The small enterprise segment is also identified as the most promising segment for growth, by both MFIs and banks. iii. Risks and obstacles to SME finance: SME-specific factors and macroeconomic factors were indicated by all MFIs and banks as significant obstacles to the development of SME lending. While banks and MFIs believe there is high growth potential in lending for small enterprises, the current lack of involvement is due to perceived risks. a. Regarding the SME specific factors, most of the financial institutions highlighted the poor quality of financial statements, inability to manage risk, lack of knowledge of business management, lack of awareness on how to be bankable, lack of adequate collateral and informality of SMEs as the major challenges. Regarding macroeconomic aspects, inflation, tax regulation and high vulnerability of the agriculture sector were mentioned as hindrances by financial institutions. b. Most financial institutions in this study perceive costs and risks to be higher in the SME segment compared to the large enterprise segment. Banks seem to have a more negative perception of risks and costs than MFIs. Further, when asked to compare the profitability of SME loans versus large enterprise loans, SME profitability is perceived to be considerably lower. iv. The missing middle phenomenon: SMEs represents a missing middle in the financial sector with high heterogeneity of lending patterns between MFIs and commercial banks a. Ethiopia lags behind other Sub-Saharan Africa’s countries and developing countries in terms of lending to SMEs. The share of SME lending in overall lending portfolio in Ethiopia is in fact only 7 percent, among the smallest shares in Sub-Saharan African countries as well as far below that of developing economies. b. SMEs represent a missing middle in the financial sector: Lending to SMEs is limited as MFI deposit and loan portfolios are comprised mainly by microenterprises (over 90 percent). The same is true for commercial bank portfolios, which are primarily comprised of large enterprises. Deposits and outstanding loans to SMEs typically comprise less than 10 percent of the total portfolios of MFIs or banks. This leaves a considerable missing-middle of SMEs not served by either banks or MFIs and who need access to finance c. High heterogeneity of lending patters: MFIs issue the most number of loans to SMEs, but banks issue the most value. Seventy-four percent of the value of SME loans outstanding from the sample of financial institutions are by banks, only a quarter of SME loan values originate from MFIs. In terms of the number of loans, ACSI has 8,670 loans outstanding to SMEs compared to only 88 by the CBE (Dec. 2012). Banks provide large loans with longer maturities compared to much smaller size lending with short maturity provided by MFIs. Interest rates for SME customers vary between 10 percent for low risk MFI customers and 15.5 percent for high risk bank customers. SMEs face higher collateral requirements and interest rates because banks perceive them as more uncertain and harder to evaluate. v. Business models. The business models of the surveyed financial institutions are mostly inadequate to serve SMEs: they lack a dedicated and specialized SME unit or department within their organizational structure; loan appraisal techniques are still mostly based on traditional relationship lending rather than on transactional technologies such as credit scoring etc.; products are highly standardized and there is very limited product innovation; distribution channels are still largely based on branches and long term financing needs of SMEs do not seem to be properly addressed SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 13 a. The organizational model used by the majority of the interviewed financial institutions does not consider the need to set up a specialized SME unit or department to better serve the SME clientele. 9 out of the 12 financial institutions did not possess a separate SME department or unit at the time of the study. b. When it comes to appraisal and monitoring of SME loans, these are largely done through the establishment of a close relationship with clients for both banks and MFIs. A minority of banks stated using transactional technologies such as credit scoring, risk rating tools, factoring or leasing. None of the MFIs use these techniques. c. The product mix offered by financial institutions does not seem to be sufficiently large. Other lending products such as leasing and factoring are not offered by any of the financial institutions in this study. Furthermore, products provided to the SME and micro enterprise market are largely standardized and efforts to continuously adapt them to client’s needs are limited. The great majority of financial institutions in the study reported no change in their financial product offering between the years 2010 and 2012. d. Geographic location does not appear to be an important marketing criterion for financial institutions. This is particularly the case for MFIs as the five dominant ones, with more than 90 percent of total asset portfolio of the micro financing sector, are affiliated with the regional governments in Ethiopia. Regarding their geographic outreach most of the banks and MFIs use only their own branches as distribution channels. e. Long term financing needs of SMEs do not seem to be properly addressed. The average loan maturity for SME loans reported by MFIs was 2.38 years while for the banks it was 6 years. According to the questionnaire responses, the average maturity loan for large enterprises was 10.4 years. This indicates that long term financing needs of SME’s are not met and that there is a potential market gap here. vi. The importance of Government financed programs: Government financed programs (credit guarantee programs and line of credit with technical assistance) are important drivers. a. When asked about the impact of government financed programs on the decision to engage in SME finance the picture that emerges clearly indicates that both categories (i.e. banks and MFIs) have a positive perception of partial credit guarantee schemes and the provision of dedicated credit lines associated with technical assistance. b. Directed credit programs are also perceived as having a positive impact, confirming, once again, the dominant role that public institutions play in the banking and microfinance sector. vii. A supportive financial sector infrastructure. The potentialities of the credit bureau are not effectively exploited. a. The survey indicated that most banks use the credit bureau information for SMEs loan analysis and most of them consider it to be an effective tool. However only the negative information provided is used and the credit bureau information has limited input and an insignificant contribution to loan decisions. b. The credit bureau’s coverage and the quality of the financial information on SMEs reported by banks should be further improved through a better enforcement of financial institutions reporting requirements. c. The fact that MFIs are still not connected to the credit bureau (despite the technological platform currently in place is capable also to accommodate MFIs) suggests the need to carefully look into the obstacles preventing MFIs to contribute to and benefit from the credit bureau service. 14 viii. Regulatory and judicial issues. The contractual environment and lack of collateral registry inhibit secured lending and constrain access to finance for SMEs. Changes in the market due to imposed lending restrictions are seen as a hindrance causing liquidity constraints. a. Lack of contract enforcement and judiciary inefficiency were indicated as main obstacles concerning the contractual environment. There is no legally authorized body to register machinery and/or equipment for it to be held as collateral. The absence of a collateral registry in combination with ineffective enforcement of contracts in case of default could contribute to significant losses for banks and thereby have significant impact on access to finance for SMEs. b. The regulatory framework affecting the liquidity position of banks is seen as another significant obstacle to involvement with SMEs. The majority of surveyed financial institutions reported that there have been significant changes in the market for lending to SMEs which affected banks in terms of liquidity and overall competition in the banking sector13. c. MFIs are more likely than banks to perceive competition in SME lending as an obstacle. Banks are more likely to perceive bank-specific factors and characteristics of SME lending as obstacles. Bank specific factors include amongst others the lack of interest at the bank level, limited geographic outreach, lack of appropriate products and knowledge on how to evaluate MSMEs or high collateral requirements. The remainder of the report is organized as follows. Chapter 2 provides a short overview of the government’s economic development goals and a summary of the Ethiopian legal and regulatory framework as well as the status of the country’s overall financial sector. Results from demand and supply-side surveys are presented and discussed in Chapters 3 and 4. Chapter 5 examines international best practices in SME finance and current government programs in support of MSMEs. Policy recommendations are presented in Chapter 6. 13 Banks in particular referred to the April 2011 NBE directive requiring private commercial banks to hold 27 percent of new loan disbursement in low-yield NBE bills. SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 15 2. MSMEs IN ETHIOPIA: THE MACRO ECONOMIC CONTEXT AND INSTITUTIONAL FRAMEWORK 2.1 Strategic and macroeconomic context for MSME GROWTH Ethiopia is currently implementing a five-year (2010/11-2014/15) Growth and Transformation Plan (GTP) in line with its long-term vision of achieving rapid, sustainable and equitable socioeconomic growth and development, reducing poverty, and meeting the Millennium Development Goals (MDGs) within the framework of macroeconomic stability. In order to promote the MSME sector the Government intends to facilitate the development of industrial clusters, for interconnected firms in a particular field with links to related institutions such as financial service providers, technical and vocational educational institutions, and various levels of government institutions. These initiatives aim to enable firms to overcome constraints in the areas of capital, skills, technology, logistics as well as to grow and compete by fostering production value chains and achieving efficiency gains. According to the Micro and Small Enterprises Development Strategy of the GTP, MSMEs are given strategic focus by the Government due to the fact that they play an essential role in the country’s industrial development plan and in the creation of employment opportunities in urban centers with the ambitious objective of 3.4 million jobs expected to be added to this segment over the five-year GTP period and training provided to over 3 million operators. Over the past decade, Ethiopia has achieved high economic growth, averaging 10.7 percent per year (World Bank, 2013) establishing the country among the fastest growing economies both in Africa and the developing world. This growth momentum has continued through the first two years of the GTP period and inflation has declined to single digits (from its peak of 40 percent in July 2011 to around 7 percent in June 2013), significantly reducing the distortionary effects that a negative real interest rate environment was having on financial intermediation. The combination of robust economic growth and increased expenditure on social assistance have resulted in a several fold rise in real per capita GDP and a dramatic drop in the national poverty rate (from 60.5 percent in 2005 to 30.7 percent in 2011)14. Economic growth has also helped to reduce unemployment particularly in urban areas. While overall unemployment rate stands at about 25 percent, urban unemployment rate declined from 22.9 percent in 2004 to 17.5 percent in 2012. In urban areas, both female and male unemployment declined and the large gap between the two groups has narrowed for the two years of available and comparable data (2009 and 2010). Unemployment rates remains particularly high among young females, with almost one third of them unable to find a job in the urban areas. For the country to continue its historically impressive growth performance (and reach the middle income status by 2025 as indicated in the GTP), the conditions for an increased scope of the private sector should be created. In fact, despite the impressive results achieved through the public sector-led growth strategy centered on high public investments promoted by the Government, the challenge of strengthening the competitiveness of the economy requires the support of the private sector as an additional engine of growth. The economic growth is driven by services and agriculture on the supply side. The share of industry in GDP has remained modest increasing to 11 percent in 2011/12 from 10.5 percent last year and, according to the Ministry of Finance and Economic Development (MoFED), its growth rate averaged 14.3 percent during the last two years (2010/11-2011/12). The modest share and growth of the industry sector and the significant share of the agriculture (44 percent) and services (45 percent) sectors clearly indicates the challenge and need for further structural transformation of the Ethiopian economy for the country to be able to join the club of middle income countries within the planned period. 14 World Bank, WDI, August 2013 SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 17 Table 1. Sectoral Contribution to GDP and GDP Growth (in Million Birr) Fiscal Year Items 2007/08 2008/09 2009/10 2010/11 2011/2012 Agriculture 170,300 181,200 195,000 212,600 223,100 38,800 Sector Industry 35,400 38,800 43,000 49,400 56,100 Services 143,100 163,200 184,700 207,900 230,900 Total GDP 346,700 381,700 421,800 469,900 510,151 Less FISM 2,400 2,700 2,900 3,200 3,600 Real GDP 344,300 378,900 418,900 466,600 506,533 Growth in Real GDP 11.2 10.0 10.6 11.4 8.6 Real GDP per capita 4,596.8 4933.6 5,316.0 5,760.5 6,088 Mid-year population 74.9 76.8 78.8 81.0 83.2 Agriculture 49.1 47.5 46.2 45.2 43.7 Share in GDP (in % ) Industry 10.2 10.2 10.2 10.5 11.0 Services 41.3 42.8 43.8 44.2 45.3 Growth in Real GDP per capita 8.1 7.5 7.3 7.8 8.4 Absolute Growth 7.4 6.4 7.6 9.0 4.9 Contribution to GDP 3.7 3.0 3.5 Agriculture growth Contribution in% 32.7 30.2 33.4 Absolute Growth 10.3 9.6 10.8 14.9 13.6 Contribution to GDP 1.0 1.0 1.1 Industry growth Contribution in % 9.4 9.7 10.5 Absolute Growth 16.1 14.0 13.2 12.6 11.1 Contribution to GDP 6.6 6.0 5.8 Services growth Contribution in% 59.3 59.8 54.6 48.8 58.4 Source: Ministry of Finance and Economic Development (MoFED) Note: Sectoral contributions do not add-up to overall GDP growth because of Financial Intermediary Service Indirect Measurement (FISIM). 2.2 Legal and Regulatory Framework The regulatory and legal environment plays an important role for the development and promotion of MSMEs and their financing needs. Financial institutions in Ethiopia operate under a rather conducive regulatory framework whereby prudential regulations and directives are in place to ensure financial system stability and financial sector soundness. The legal and regulatory environment is conducive for providing finance from banks and MFIs to business activities. Particularly, micro finance business proclamation number 626 gives flexibility for MFIs to transform into banks or any other financial institutions that they would be able to provide financial services to small and medium size enterprises. The regulatory framework has been relaxed on maximum loan size from time to time. National Bank of Ethiopia (NBE) directive number MFI/18/06 provides extension to any single 18 borrower up to 1 percent of the total capital of the MFI which is good enough for MFIs to lend to SMEs. However, the directive sets regulatory requirements that constrain MFIs to predominately make credit available on the basis of group guarantees amongst members who have joined a membership arrangement. The directive also has a provision for lending to individuals on a limited scale on the basis of collateral. However, the absence of a well-functioning and efficient legal system to enforce contracts affects the implementation of prudential regulations in the banking and microfinance sector. In Ethiopia’s court system, commercial benches are handling the court cases related to businesses. There are no dedicated commercial units in the first instant courts to relieve the commercial cases and to strengthen contract enforcement. The framework for banking supervision and regulation was strengthened during a financial sector reform in the areas of financial reporting, external audits and related frameworks and deposit insurance. NBE, thanks to the World Bank’s Financial Sector Capacity Building Project, has now adopted risk based supervision and conducts quarterly credit and liquidity risk stress testing of the banking system. Commercial banks continue to be subject to the April 2011 NBE directive requiring them to hold 27 percent of new loan disbursements in low-yield NBE bills. The proceeds of these bonds are transferred to the state-owned Development Bank of Ethiopia (DBE) which, according to the stated policy is supposed to on-lend them to government targeted private sector activities. The application and the effects of this directive have been controversial. The Ethiopian Bankers’ Association (EBA) has appealed to NBE to revise this directive15 arguing that it forced banks to unnecessary portfolio adjustments. In fact, to avoid the effects of this 27 percent bond requirement, private banks have been moving up market and towards longer term financing. Aiming at counteracting this tendency and to push for T-bill purchases, NBE has issued a new directive in February 2013 setting the minimum requirement for short term loans at 40 percent of bank’s total loan provision. This directive applies to all banks (except DBE and CBE) and forces them to provide short term loans with maturities of less than a year. The directive also cuts commercial banks’ total deposit reserve requirements down to 5 percent, from the previous 10 percent. The aim of this directive is to counter the tendency of private banks to move up market. These aspects will be addressed further in the section on obstacles to SME finance. The directive is effective since beginning of July 2013. NBE has also given banks until January 2015 to restructure their loan portfolios to the required ratio16. On the other side, the National Bank of Ethiopia directive number FIS/01/2012 on regulation of mobile and agent banking services is a very positive step to stimulating the use of technology and innovative financial service delivery channels such as mobile devices and agents that will have significant contribution in deepening financial service accessibility to the a wider section of population at an affordable price. 2.3 Outlook of The Ethiopian Financial Sector Ethiopia’s financial sector is dominated by the banking sector (commercial banks) which currently represents more than 92.6 percent of total assets of the financial sector, excluding the assets of the Development Bank of Ethiopia (DBE) and National Bank of Ethiopia (NBE). MFIs constitute 5.2 percent and insurance companies 2.2 percent of the total financial sector assets. There is a considerable increase in assets of MFIs from 4.4 percent of the total financial sector assets in 2005/06 to 5.2 percent in 2011/12 demonstrating the increasing role of microfinance institutions in poverty alleviation, asset building and employment creation particularly in rural communities. Financial intermediation is a driving force for economic development: an expansion in credit to the private sector in fact enables firms to invest in productive capacity thereby laying the foundation for a sustainable growth path. However, Ethiopia is falling behind its peers in this area (Figure 5). In 2011, credit to the private sector in Ethiopia was equivalent to about 14 percent of GDP compared to the regional average of 23 percent of GDP. Moreover, while the worldwide trend has been an 15 http://addisfortune.com/Vol_10_No_581_Archive/Bankers%20Association%20Wants%20T-bill%20Directive%2 Revised.htm 16 http://www.capitalethiopia.com/index2.php/index.php?option=com_content&view=article&id=2543:nbe- restructures-loan-portfolio-reduces-reserve-requirement&catid=54:news&Itemid=27 http://www.bloomberg.com/news/2013-02-26/ethiopian-central-bank-order-may-mean-more-t-bills-less- lending.html October 2012 - IMF Country Report No. 12/287 SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 19 There is a considerable increase in assets of MFIs from 4.4 percent of the total financial sector assets in 2005/06 to 5.2 percent in 2011/12 demonstrating the increasing role of microfinance institutions in poverty alleviation, asset building and employment creation particularly in rural communities. Financial intermediation is a driving force for economic development: an expansion in credit to the private sector in fact enables firms to invest in productive capacity thereby laying the foundation for a sustainable growth path. However, Ethiopia is falling behind its peers in this area (Figure 5). In 2011, credit to the private sector in Ethiopia was equivalent to about 14 increase in private percent of GDPsector credit, to compared Ethiopia average of a23 has experienced the regional decline of about percent 5 percentage Moreover, points of GDP. while the since 2004. worldwide trend has been an increase in private sector credit, Ethiopia has experienced a decline of about 5 percentage points since 2004. Figure 5. Ethiopia: Private sector credit (% of GDP) Figure 5. Ethiopia: Private sector credit (% of GDP) 40 35 30 25 20 15 10 5 0 2004 2005 2006 2007 2008 2009 2010 2011 Ethiopia SSA Average LIC Average Kenya Tanzania Source: World Bank (FinStats 2012) Source: World Bank (FinStats 2012) The banking Thesector. Government-owned banking banks dominate sector. Government-owned banks the Ethiopian dominate thebanking Ethiopiansystem and system banking this makesandof this Ethiopia makes exception an of Ethiopia within Sub-Saharan an exception withinAfrica and across Sub-Saharan the developing Africa and across world, where the developing banking systems world, where have higher shares muchsystems banking have of private much and foreign higher shares of private andPublic participation. banks, foreign which participation. mainly focus Public on financing banks, which large mainly focus on are enterprises, dominating financing large the credit market enterprises, share of lending are dominating in the credit market share the banking sector.of sharein lending The of the banking private banks sector. The sharecredit in outstanding of private banks lending has in outstanding dropped credit from 39 lending percent of the has dropped market from share in 39 percent 2009/10 to of 32the market percent in share 2011/12 in 2009/10 while thatto of 32thepercent publicinbanks 2011/12 rose while from 61 that 68 the to of public percent banks during rose the samefrom 61 to period, 68 percent 2009/10 during the to 2011/12. same Table period, that 2009/10 2 indicates the to total2011/12. disbursementTableof indicates 2 public thathas banks total disbursement the almost tripled in the of public last three banks yearshas almost during tripled in 2009-2012 as the last (particularly public banks three years the Commercial during 2009-2012 Bankasof Ethiopia) public banksfocused on financing (particularly thelarge scale public Commercial Bank of infrastructure projects. Ethiopia) focused on same At the time, financing large scale29 the lending capacity public of Development infrastructure projects.Bank of Ethiopia At the same time, the was enhanced lending through capacity the introduction of Development of NBE Bank bills of while the was Ethiopia annual new credit enhanced disbursement through of the introduction of NBE has private banks increased bills while the by only 28 annual percent new creditin the same period. disbursement of private banks has increased by only 28 percent in the same period. Table 2. Loans and Advances by Lenders (in Million Birr) Table 2. Loans and Advances by Lenders (in Million Birr) 2009/10 2010/11 2011/12 Outstanding Outstanding Outstanding Increment in Lenders Disbursement Collection Credit Disbursement Collection Credit Disbursement Collection Credit disbursement A. Public Banks - amount 13,939.30 10,168.00 33,394.60 21,955.80 11,987.80 47,924.80 36,949.20 18,479.90 75,250.10 165% - percent 48% 41% 61% 52% 39% 65% 66% 53% 68% B. Private Banks - amount 14,965.80 14,898.80 21,297.50 20,252.00 18,560.40 26,046.60 19,152.90 16,707.50 34,950.50 28% - percent 52% 59% 39% 48% 61% 35% 34% 47% 32% Total 28,905.10 25,066.80 54,692.10 42,207.90 30,548.20 73,971.40 56,102.10 35,187.40 110,200.60 Source: NBE reports Note: Source: NBE Outstanding Credit excludes lending to central government. reports Note: Outstanding Credit excludes lending to central government. Despite the overall disintermediation trend, the Ethiopian financial sector continues to Despite the overall disintermediation trend, the Ethiopian financial sector continues to have the have the potential to be a driver of growth. The banking sector remains stable, well-capitalized to be potentialand a driver of continues togrowth. The profitable. be highly banking sector remains Figure 6 showsstable, how well-capitalized the Ethiopian and continues banking sector ranks to be highly profitable. Figure 6 shows how the Ethiopian banking sector ranks higher higher than the SSA average in terms of profitability measured on the basis of Return than the SSA on Equity average in of profitability termsHigh (ROE). measured profitability on the basis is also explained byof Return limited on Equity (ROE). competition. High profitability Although the total number of is also explained by limited banks operating incompetition. Ethiopia has Although total number thefrom increased of banks 11 in 2006 operating to 18 in 2012, in the Ethiopia bank assets has increased from 11 in 2006 to 18 in 2012, the bank assets concentration index concentration index (focusing on the 3 biggest banks) shows that Ethiopia’s banking (focusing on thesector is 3 biggest banks) much shows more that Ethiopia’s concentrated than banking the SSA sector and Lowis much more Income concentrated Group averages. than the SSA The non-performing and LowloanIncome ratioGroup averages. continued The non-performing its downward trend reaching an ratio loan continued unusually its downward low value trend of 1.4 percent in 2013. NBE attributes this improvement in the NPL ratio to effective action plans that commercial banks adopted following an NBE directive as well as to the impact of closer monitoring. 20 reaching an unusually low value of 1.4 percent in 2013. NBE attributes this improvement in the NPL ratio to effective action plans that commercial banks adopted following an NBE directive as well as to the impact of closer monitoring. Figure 6. Ethiopia: profitability of the banking sector (ROE) Figure 6. Ethiopia: profitability of the banking sector (ROE) 50.0 Ethiopia 45.0 Kenya 40.0 SSA Average 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Source: World Bank (FinStats 2012) Source: World Bank (FinStats 2012) The microfinance sector. Microfinance is a dynamically developing sector in the financial industry The in Ethiopia. Atmicrofinance the end of 2012, sector. 30 licensedis Microfinance there were a dynamically microfinance developing institutions (MFIs) sector operatingin the financial industry in Ethiopia. At the end of 2012, there were 30 licensed in Ethiopia. Their deposits amounted to Birr 5.5 billion and represented the savings of 2.6 million microfinance institutions clients. (MFIs) Some MFIs operating are sizeable in Ethiopia. financial Their institutions deposits in their amounted own rights to Birr and bigger than5.5 some of the and billion represented the savings of 2.6 million commercial banks. The sector is highly concentrated, with the 5 largest MFIs (owned by regional in clients. Some MFIs are sizeable financial institutions their own rights governments and bigger and operating some without than regions in different commercial of the competing withbanks. The corresponding sector is highly each other) concentrated, to 89% of total with sectorthe 5 largest assets and 83%MFIs of(owned by regional total borrowers. The governments predominant and loan operating in different methodology is regions group without loans, competing but some each other) with started MFIs have corresponding to offer to 89% individual loans. Thisof total sector needs development assets to andbe83% of total borrowers. expanded to support theThe predominant development loan methodology of MSMEs. MFIs are not yetis group loans, exchanging but some through information MFIs have started the credit to individual offer at bureau the National loans. Bank ofThis development Ethiopia, althoughneeds to bureau’s the credit be expanded to support would the technology development allow for it. of MSMEs. MFIs are not yet exchanging information through the credit bureau at the National Bank of Ethiopia, although the credit bureau’s technology would allow for it. Access to financial services. Access to financial services remains highly limited all over Ethiopia with only 1.97 commercial bank branches and 0.33 ATM per 100,000 adults (compared for instance Access to financial services. Access to financial services remains highly limited all over to Kenya where there are 5.17 commercial branches and 9.46 ATMs per 100,000 adults). Ethiopia with only 1.97 commercial bank branches and 0.33 ATM per 100,000 adults (compared for instance Access to Kenya to finance a top there where remains are for obstacle 5.17 commercial enterprises. Thebranches and 9.46 data recently published ATMs per 2011 of the 100,000 adults). Ethiopia Enterprise Survey (WB, 2012) confirm that access to finance remains a top obstacle for enterprises: this is perceived as the main business environment constraint by micro (41%), small (36%), andAccess to finance medium top remains a in (29%) enterprises obstacle Ethiopia, for enterprises. compared to an SSAThe recently average published of 24%, data of 20%, and the 2011 Ethiopia Enterprise Survey (WB, 2012) confirm that access to finance 16% respectively (Figure 7). The same survey indicates that almost 93% of small enterprises andremains a top obstacle for enterprises: this is perceived as the main business environment constraint over 95% of medium enterprises have either a checking or a savings account (a percentage higher by micro (41%), than the small (36%), respective SSAand medium averages) but(29%) enterprises only 3% in Ethiopia, of small enterprises andcompared to an SSA 23% of medium haveaverage a loan of 24%, 20%, and 16% respectively (Figure 7). The same survey indicates that almost 93% or a line of credit. These low percentages can be explained by (among other factors) the extremely of small enterprises and over 95% of medium enterprises have either a checking or a savings account (a percentage higher than the respective SSA averages) but only 3% of small enterprises and 23% of medium have a loan or a line of credit. These low percentages can be explained by (among other factors) the extremely high value of collateral needed for a loan, corresponding to 249.3% (253.5%) of the loan amount for small (medium) enterprises, against a SSA average of 160%. 31 SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 21 high value of collateral needed for a loan, corresponding to 249.3% (253.5%) of the loan amount for small (medium) enterprises, against a SSA average of 160%. This also means that there is a very This potential high also meansfor there is cross-selling profitable that to be exploited a very high potential by banks for profitable if they target cross-selling these to be existing exploited by account holders and offer them credit. banks if they target these existing account holders and offer them credit. Figure 7. Ethiopia: Access to Finance is a Top Obstacle Figure 7. Ethiopia: Access to Finance is a Top Obstacle 50 40 % of Firms 30 20 10 0 Ethiopia 2011 SSA World Access to Finance Micro Small Medium (21-99) Large (100+) EnterpriseSurveys Source: Enterprise Source: Surveys The financial sector infrastructure. The financial sector capacity building program conducted The financial from 2006/7 to 2010/11 sector infrastructure. with the technical and The financial financial supportsector capacity of the World building Bank has program brought conducted from 2006/7 to 2010/11 with the technical and financial significant improvements to soundness and enhanced business transactions and support of the World of outreach Bank the has brought significant improvements to soundness and enhanced business transactions sector. The major contribution of the capacity building program was introducing key financial sector and sector. The outreach of theincluding infrastructures major contribution a modern national paymentof the system capacity building and program was credit information introducing system which key financial sector infrastructures including are currently in operation to a certain extent. a modern national payment system and credit information system which are currently in operation to a certain extent. However, at present the payment system provides only large value transactions and cheques clearing retail at However, system while the payment presentinstruments payment and system provides innovative productsonly large need valueexpansion further transactions and in order cheques clearing system while retail payment instruments and innovative products to provide efficient payment services to businesses. A central switch system is not yet in place and need further expansion most of thein banks efficient order to provide commercial are still inpayment the processservices to businesses. of completing theirA central core switch banking system systems. is not yet in place and most of the commercial banks are still in the process of completing their core the Also credit banking information system is limited to a public credit registry. There is no private credit systems. bureau that would provide value added services on credit information such as credit scoring, marketingAlso the credit service, information application system processing is portfolio and limited to a public credit monitoring registry. to facilitate There quality is no decision private on making credit creditbureau would that by provision provide financial value added institutions services and thereby on credit minimize information fraud and mitigatesuch as risks. credit scoring, marketing service, application processing and portfolio monitoring to facilitate quality decision making on credit provision by financial institutions and thereby minimize fraud and mitigate risks. 32 22 3. DEMAND-SIDE ANALYSIS Main research questions addressed: 1. Which firms create more jobs in Ethiopia? 2. Which firms are the most credit constraint? 3. Is there a link between access to finance and firms’ performance? This section utilizes two data sets to describe the characteristics of Ethiopian MSMEs vis-à-vis other companies with a special focus on identifying their constraints in accessing external finance. The focus of this chapter is on three inter-related aspects: job creation, access to finance, and firm performance. The two main sources of data used for the analysis are: 1. The Ethiopia Survey of Large and Medium Scale Manufacturing Industries (LMMIS): an unbalanced panel composed of about 6,000 firms with at least 10 employees which allows for a study of firm dynamics from 2000 through 2011. 2. The World Bank’s Enterprise Survey (ES): conducted between July 2011 and July 2012 and including 794 firms which allows for the additional examination of the services sector, microenterprises, and a more detailed understanding of firm experiences with respect to access to finance. These two datasets offer complementary information and reveal supporting trends17. The main findings from the demand-side analysis are summarized below and then discussed in-depth in the following sub-sections. i. Job creation and employment growth a. In both service and manufacturing sectors, job creation is concentrated in large established (i.e. older) firms. Jobs analysis in the manufacturing sector shows that the majority of paid employment is found in large enterprises. Interestingly, in the services sector the job dynamics are similar to the ones in the manufacturing sector as large established firms tend to be the most important net jobs creators during the period 2009-2011. b. Typically, young firms are a great source of job creation, but this trend is not seen in Ethiopia. Conditional on survival, young Ethiopian firms enjoy higher employment growth rates than established firms, but they also generate a smaller number of new jobs than older firms. In the rest of the world, young firms are principal job creators despite high exit rates (Haltinwanger et al 2011; Rijkers et al, 2013). The fact that older firms dominate net job creation in Ethiopia is worrisome since it suggests there is a lack of competitiveness and innovation in the private sector. c. Job creation and employment in the services and retail sectors are higher than in the manufacturing sector. The manufacturing sector plays a limited role in the overall Ethiopian economy, comprising only 4.2 percent of GDP in 2012/13. Therefore, it’s important to also understand employment trends in the services sector. The service sector created nine times more jobs than the manufacturing sector during the period 2009-2011. ii. Access to finance a. Overall, data indicates the existence of a missing middle phenomenon in terms of financial services catering to small firms. Only 1.9 percent of small firms have a loan or line of credit. This rate is much lower than that of micro, medium, and large firms (6.0, 20.5, and 35.5 percent respectively). 57 percent of small firms are fully credit constrained, a rate higher than in any other size group. These statistics corroborate with assertions that small firms struggle the most in obtaining access to finance since 17 More information on these two data sets can be found in Appendix 2 and 3. SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 23 MFIs cater to micro-sized firms, and commercial bank clientele are predominantly medium and large firms. b. Young and smaller firms are much more likely to be rejected for a loan or a line of credit than firms who are more established or larger. Among firms who applied for a loan or line of credit in fiscal year 2011, 57.3 and 87.9 percent of applications submitted by micro and small firms respectively were rejected; this is in sharp contrast to the much lower 6.2 and 10.4 percent rejection rate experienced by medium and large firms. 56 percent of young firms had their loan applications rejected compared to 33 percent of established firms. c. Despite confirming the need of increased access to finance, SMEs are discouraged from applying for loans due to high collateral requirements. Results show that SMEs are less likely to say that they don’t need a loan, thus confirming the need for finance. However, SMEs are discouraged (or voluntarily exclude themselves) from applying for loans or lines of credit. About a third of SMEs reported that they did not apply for a loan or line of credit because collateral requirements were too high. Collateral rates in Ethiopia are much higher than in more developed African economies. For example, collateral rates are only 120.8 percent of the loan value in Kenya (2007) compared to 234 percent in Ethiopia. Small firms are also the most likely to use personal assets as a type of collateral (36.8 percent of small firms use this type). iii. Firms’ performance a. Firms that are credit constrained exhibit poorer performance and productivity. In Ethiopia, a firm that is credit constrained has sales growth that is 15 percentage points lower, employment growth that is 5 percentage points lower, and labor productivity growth that is 11 percentage points lower than firms who are not credit constrained. b. Investment decisions of manufacturing firms in Ethiopia are heavily dependent on cash flows. To identify the existence of credit constraints we evaluate the extent to which firms’ investments are reliant on cash flow. We find some evidence that liquidity constraints are stronger for young small firms when compared to other small firms. On the other hand, there is limited evidence that young medium and large firms are more credit constrained than other medium and large firms. 3.1. The place of small and medium enterprises in the economy What is the role and place of SMEs in terms of job creation? We first examine employment trends in the manufacturing sector between 2000 and 2011 using the Ethiopia Survey of Large and Medium Scale Manufacturing Industries (LMMIS). The data set is an unbalanced panel comprised of 495 – 1,039 firms per year18. Our results, shown in Figure 8, suggest that medium and large firms are creating most of the new jobs during this period19. The top panel illustrates the total number of employees in the manufacturing sector, and the lower panel illustrates the number of net new employees disaggregated by firm size based on Ethiopia’s Central Statistical Agecny definitions (micro (<10); small (10-19), medium (20-99), large (100+)), as well as firm age (young (0-5 years); established (6+ years)). 18 An important note about this data set is that this survey is designed to exclude small-scale industries (microenterprises) with fewer than 10 employees however some firms in the LMMIS data set reported fewer than 10 workers. Employment levels calculated using this survey will be negatively biased for microenterprises. The micro segment of the manufacturing sector is extremely large and important. In fiscal year 2007/08, the Central Statistical Agency reported there were 43,338 microenterprises in Ethiopia comprising of 138,951 engaged persons. In the LMMIS, we have only 35 microenterprises in fiscal year 2007. 19 These statistics do not capture changes in part-time employment or shifts in hiring from full-time to part-time employees. By construction these statistics only consider firms who were in operation in 2009, therefore new jobs created by firms who began operations after 2009 is not captured. 24 It is also important to distinguish among different types of employees. The results show that paid employees are primarily employed in large establishments, while the composition of workers in smaller scale enterprises tend to be characterized by a higher prevalence of temporary and unpaid workers, relative to large establishments. Therefore, when taking into account all workers (“engaged persons”) including paid employees, unpaid working proprietors, active partners, unpaid family workers and paid and unpaid apprentices, the weight of MSMEs increases. For example, in 2011 half all engaged persons were affiliated with MSMEs. Large manufacturing companies that have been established for more than 5 years have been the largest employers of permanent paid workers in the economy during the last decade. In 2011, they accounted for 74 percent of total paid employment. Another important result is relative to the effect of exogenous shocks on job creation, revealing the sensitivity of the Ethiopian manufacturing sector to external shocks. Specifically, in the aftermath of the financial crisis the most affected firms were large and established ones. In 2008, these firms experienced a net loss of 4,043 jobs. However, this was just a temporary shock as, in the next year, jobs began to recover slowly. In 2009, the large enterprise saw an addition of 381 jobs and 5,480 jobs in 2010. Another way to examine employment trends is to look at the proportion of firms in a size group in year t who transition to another size group in year t+1 (Table 3). In the table below, we examine firm size transitions only in recent years after the financial crisis (t=2008, 2009, and 2010). Large firms are the most stable, the majority of large firms stay large, and virtually none reduce in size to less than 20 employees. Micro firms are the most dynamic, over half of micro firms expand to become small, medium, or large firms in the next year. Small firms are much less dynamic, as only 11.1 percent of small firms grow to become medium firms in the next year. Growth for medium firms is also challenging after the financial crisis; medium firms are more likely to become smaller (11.6 percent) than to become larger (4.5 percent). Table 3. Firm Size Transition in Manufacturing (t=2008-2010) Year t+1 Micro Small Medium Large (0-9) (10-20) (21-99) (100+) Micro (0-9) 49.8% 43.8% 6.0% 0.5% 100% Small (10-20) 10.3% 78.1% 11.1% 0.5% 100% Year t Medium (21-99) 1.7% 9.9% 84.0% 4.5% 100% Large (100+) 0.8% 0.6% 7.5% 91.1% 100% Source: Authors calculation based on data from Ethiopia’s Survey of Large and Medium Scale Manufacturing Industries (2008 - 2011) (LMMIS). This survey is designed to exclude small-scale industries with fewer than 10 employees however some firms in the LMMIS data set reported fewer than 10 workers. SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 25 Figure 8. Employment Trends in Manufacturing (2000-2011), by Age and Size Figure 8. Employment Trends in Manufacturing (2000-2011), by Age and Size Groups Groups Young (0-5 Years) Established (6+ Years) 120000 100000 Total Number of Employees 80000 60000 40000 20000 0 2000 2002 2004 2006 2008 2010 2001 2003 2005 2007 2009 2011 Micro Small Medium (21-99) Large (100+) Net  Job  CreaIon  from  2000-­‐2011,  by  segment   60000   50000   40000   30000   20000   10000   0   Small  (10-­‐20)   Medium  (21-­‐99)   Large  (100+)   Established  (6+  years)   Young  (0-­‐5  years)   Source: Authors calculation based on data from Ethiopia’s Survey of Large and Medium Scale Source: Authors calculation Manufacturing based Industries on data (2000 from - 2011) Ethiopia’s (LMMIS). Survey This of Large survey and Medium is designed Scale to exclude Manufacturing small-scale industries Industrieswith fewer (2000 10 employees than(LMMIS). - 2011) however This survey to in some firms is designed the LMMIS exclude data set small-scale reported with industries fewer than fewer 10 workers. than 10 employees however some firms in the LMMIS data set reported fewer than 10 workers. However, it is important to widen our analysis to the entire economy, since the large and medium scale manufacturing sector contributed to only 2.9 percent of GDP in 2012/13, while services accounted for 45.2 percent of GDP. An understanding of employment dynamics in the services sector is important. 37 To compare employment trends between manufacturing and services, we turn to the World Bank’s Enterprise Surveys (ES). The Enterprise Surveys interviewed 794 Ethiopian firms in 2011 that were formally registered, had private ownership, and operated in the manufacturing and service/retail sectors. The Enterprise Survey is representative of the entire population of Ethiopian firms at the level of sector, size, and sub-national region. Another important feature of the Enterprise Surveys is the inclusion of microenterprises since these firms are not targeted in Ethiopia’s Survey of Large and Medium Scale Manufacturing Industries; 154 microenterprises were surveyed in the Addis Ababa region only. However, while it is now possible to compare the manufacturing section with 26 services, and also analyze microenterprises, we lose the ability to follow trends before and after the financial crisis. Analyzing the ES we find that, conditional on survival, the majority of Ethiopian firms were job creators. Less than 8 percent of firms reported employing fewer full-time permanent workers in fiscal year 2011 than in 2009 (Figure 9)20. However, a significantly larger number of jobs were created in the services and retail sector compared to the manufacturing sector. Less than 10 percent of all new jobs were created in the manufacturing sector. During the period 2009-2011, medium and large firms created most of the new jobs.21 Established firms (i.e. firms that have been in operations for six years or more) added 15,615 full-time permanent jobs, equivalent to 38 percent of all new jobs. Large firms who are young created the smallest number net jobs. Small established firms that remain small after many years in operations are likely to be poor performers, and accordingly created fewer jobs than young small firms. Large established firms create the most number of jobs; however, younger firms have the highest growth rates (Figure 10). In Ethiopia, young firms do not contribute to job creation as significantly as in the rest of the world. In the U.S. younger firms grow faster than more established firms, and also yield higher net job creation (Haltiwanger et al 2011). Rijkers et al (2013) also finds that young firms in Tunisia create the most new jobs despite high exit rates. This shows that Ethiopian firms are not competitive and it is difficult for new start-ups to survive. Figure 9. Net New Jobs (2009-2011), by Industry Figure 9. Net New Jobs (2009-2011), by Industry Manufacturing  Only   1500   1056   1035   Net  New  Jobs   1000   793   500   305   300   271   0   Small  (10-­‐20)   Medium  (21-­‐99)   Large  (100+)   Age=1-­‐5   Age=6+   Retail  &  Services  Sectors   20000   14581   Net  New  Jobs   10000   6276   3420   4122   4282   982   0   Small  (10-­‐20)   Medium  (21-­‐99)   Large  (100+)   Age=1-­‐5   Age=6+   Figure 10. Annual Employment Growth Rate (%) (2009-2011), by Industry Manufacturing  Only   20   14.8   16.2   Annual  Employment   8.6   6.5   7.8   Growth  Rate  (%)   10   2.7   0   Small  (10-­‐20)   Medium  (21-­‐99)   Large  (100+)   20 These estimates are computed from surviving firms and may be positively biased since the Enterprise Surveys excludes firms who closed. Age=1-­‐5   Age=6+   21 These statistics do not capture changes in part-time employment or shifts in hiring from full-time to part-time employees. By construction these statistics only consider firms who were in operation in 2009, therefore new jobs Retail  &  Services  Only   created by firms who began operations after 2009 is not captured. 17.2   20   13.0   10.6   mployment   9.5   7.2   8.5   h  Rate  (%)   10   SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 27 0   Small  (10-­‐20)   Medium  (21-­‐99)   Large  (100+)   20000   14581   Net  New  Jobs   10000   6276   3420   4122   4282   982   0   Small  (10-­‐20)   Medium  (21-­‐99)   Large  (100+)   Age=1-­‐5   Age=6+   Figure 10. Annual Employment Growth Rate (%) (2009-2011), by Industry Figure 10. Annual Employment Growth Rate (%) (2009-2011), by Industry Manufacturing  Only   20   14.8   16.2   Annual  Employment   8.6   6.5   7.8   Growth  Rate  (%)   10   2.7   0   Small  (10-­‐20)   Medium  (21-­‐99)   Large  (100+)   Age=1-­‐5   Age=6+   Retail  &  Services  Only   17.2   20   13.0   10.6   Annual  Employment   9.5   7.2   8.5   Growth  Rate  (%)   10   0   Small  (10-­‐20)   Medium  (21-­‐99)   Large  (100+)   Age=1-­‐5   Age=6+   Notes: Net new jobs is calculated as the difference in the number of full-time permanent workers in fiscal Notes: Net new jobs is calculated as the difference in the number of full-time permanent workers in fiscal years 2011 and 2009.Annual employment growth rate is calculated using the number of full-time permanentyears 2011 and employees 2009.Annual in fiscal employment years 2011 growth and 2009. Growth rate rates is calculated are calculated fromusing the firms surviving number of full-time only. permanent Source: Enterprise employees in fiscal years 2011 and 2009. Growth rates are calculated from surviving firms Surveys only. Source: Enterprise Surveys 39 3.2 Is access to finance an obstacle to business? 3.2.1 Firms’ perceptions of access to finance in Ethiopia Young and small firms appear to be facing more serious financial constraints relative to those that are larger and more established (Figure 11). Across a range of finance indicators created using the Ethiopia Enterprise Surveys (2011), young and small firms are the most likely to report that access to finance is a major constraint to their business operations, and at rates higher than other well-developed African countries. In South Africa (2010), only 10.4 percent of SMEs22 rated access to finance as a major constraint, compared to the much higher rates in Ethiopia. Nearly half of microenterprises, 40 percent of small firms, and 18.5 percent of medium firms reported access to finance in Ethiopia to be a major constraint to daily operations. Consistent with this “perception based indicator”, objective data show smaller firms are more likely to be excluded from financing and utilize fewer financial instruments. In fact, as shown in the lower panel of Figure 11, multiple objective indicators suggest that micro and small firms, as well as those that are young are excluded from the financial market. While the majority of micro and small enterprises do have checking and savings accounts, access to finance is extremely low for this segment. Only 6 percent of micro enterprises, 1.9 percent of small enterprises, and 20.5 percent of medium have a loan or a line of credit. Small firms also have the highest incidence of loan application rejection. Among firms who applied for a loan or line of credit in the last fiscal year, 57.3 and 87.9 percent of applications submitted by micro and small firms respectively were rejected. This is in sharp contrast to the much lower 6.2 and 10.4 rejection rate experienced respectively by medium and large firms. In addition, less than 1 percent of micro firms have external financing at all, whether to finance working capital or investments. Similarly, only 4.5 percent of micro firms have an overdraft facility, compared to 57.1 percent of large firms. 22 The South Africa (2007) Enterprise Survey did not survey micro firms so we only have SME and not MSME comparisons. 28 These low lending and financing rates to small firms can be attributed to (among other factors) the extremely high value of collateral needed for a loan. Figure 11. Access to Finance is a larger obstacle for young and small firms Figure 11. Access to Finance is a larger obstacle for young and small firms 60   55.9   48.8   50   40   32.6   32.1   30   25.3   20   14.0   9.9   8.4   7.3   6.7   10   0   Access  to  Finance  is  a   Loan  applicaIon  was   Has  a  Loan  or  Line  of   Has  an  overdraX  facility  Has  External  Financing   Major  Constraint   rejected   Credit   Young  (Age  0-­‐5)   Old  (Age  6+)   100   87.9   90   80   70   57.3   57.1   60   49.8   50   39.5   40.8   35.5   40   30.8   30   24.2   18.1   20.5   15.2   20   10.4   10.3   6.2   6.0   4.5   10   1.9   0.5   3.5   0   Access  to  Finance  is  a   Loan  applicaIon  was   Has  a  Loan  or  Line  of   Has  an  overdraX   Has  External  Financing   Major  Constraint   rejected   Credit   facility   Micro  (0-­‐9)   Small  (10-­‐20)   Medium  (21-­‐99)   Large  (100+)   Notes: Micro firms 0-4 employees were surveyed in Addis Ababa only. Weights are applied to the Notes: Micro firms 0-4 employees were surveyed in Addis Ababa only. Weights are applied to the calculations. calculations. Loan application refers to the most recent application that was submitted in the last fiscal year. Loan If no loan application applications wererefers to the submitted in most recent the last fiscal application that year, then this was submitted question in the last fiscal year. If no was skipped. loan applications Source: Enterprise Surveys were submitted in the last fiscal year, then this question was skipped. Source: Enterprise Surveys A financial environment in which firms, especially small ones, face high barriers to obtaining finance leads A financial environment to voluntarily in which exclusion firms, especially or discouragement small from applying ones, face high barriers for a finance to obtaining loan. In fact, as shown in Table 4, when firms were asked why they did not leads to voluntarily exclusion or discouragement from applying for a loan. In fact, as shown apply for a in loan, only a small minority of small firms respond there was no need for a loan Table 4, when firms were asked why they did not apply for a loan, only a small minority of small (34% compared firms respondto 62%thereof large was firms). no need In for comparison a loan in South to (34% compared Africa, 62% ofabout large80 percent firms). of In comparison small23, medium, and large firms who didn’t apply for a loan reported it was because in South Africa, about 80 percent of small , medium, and large firms who didn’t apply for a loan 23 there was no need for a loan rather than other reasons (Fuchs et al, 2011). Again, this is reported it was because there was no need for a loan rather than other reasons (Fuchs et al, 2011). consistent with the idea of a missing middle phenomenon hindering opportunities to Again, this is consistent with the idea of a missing middle phenomenon hindering opportunities access finance for small companies, but not micro or large enterprises. Microenterprises to access finance for small companies, but not micro or large enterprises. Microenterprises are less are less likely to report that high collateral or interest rates were reasons for not applying likely to report that high collateral or interest rates were reasons for not applying for a loan than for a loan than SMEs. This reflects that MFIs appear to be doing a good job serving SMEs. This reflects that MFIs appear to be doing a good job serving microenterprises; on the other hand 23 small firms do not seem to be able to access MFIs while at the same time being underserved No micro firms were surveyed in South Africa by the banking sector. 41 23 No micro firms were surveyed in South Africa SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 29 Table 4. Reasons for Not Applying For a Loan (% of Firms by Size Group) Micro Small Medium Large (0-9) (10-20) (21-99) (100+) No need for a loan - sufficient capital 48% 34% 45% 62% Application procedures were complex 9% 6% 5% 7% Interest rates were not favorable 4% 10% 4% 6% Collateral requirements were too high 20% 29% 33% 6% Size of loan and maturity were insufficient 0% 2% 1% 2% Did not think it would be approved 13% 6% 4% 4% Other 5% 13% 8% 13% Total 100% 100% 100% 100% Notes: This question was asked to firms who did not apply for a loan or line of credit in fiscal year 2011. Source: Enterprise Surveys It is clear why collateral requirements can be binding for smaller firms since the most common type of collateral used are land and buildings and personal assets (Table 5). Like elsewhere in developing economies, Ethiopian banks prefer immovable collateral such as land rather than movable assets such as machinery. Large firms are the only ones who commonly can use equipment as collateral. The use of accounts as collateral is also rare, even for large firms, less than a quarter of large firms use this as a form of collateral. The average value of collateral needed for loans in Ethiopia is also very high compared to other regions of the world as well as to other developed economies in Africa. On average, Ethiopian firms require 234 percent of the loan amount for collateral, compared to 134.3 percent in Eastern Europe and Central Asia. In well-developed African countries, collateral requirements are also much lower than in Ethiopia: 120.8 percent in Kenya (2007), and 103.6 percent in South Africa (2007). Table 5. Types of Collateral Used Micro Small Medium Large (0-9) (10-20) (21-99) (100+) Land and Buildings 69.6 86.1 81.9 85.4 Equipment 2.1 2.5 33.0 84.9 Accounts 2.1 2.5 4.8 24.5 Personal Assets 26.2 36.8 27.0 22.0 Other 4.2 0.0 0.0 14.3 Notes: This question was asked to firms who received a loan or line of credit in fiscal year 2011. Firms may cite the use of more than one type of collateral. Source: Enterprise Surveys These indicators illustrate a clear picture that in Ethiopia, young and small firms are less likely to utilize lending services and also face barriers when they attempt to acquire financing. 3.2.2 More than elsewhere firms are considerably credit constrained Figure 1 illustrates the percentage of firms that are credit constrained using a measure that distinguishes between four degrees of “credit constrained”24 which is constructed using the Ethiopia Enterprise Surveys (2011). The main result is that, while the proportion of firms in Ethiopia who are not credit constrained (NCC) is comparable to the Sub-Saharan Africa regional average, the share 24 See Appendix 4 for full definition. 30 of firms completely without access is much higher. In other words, firms in Ethiopia are much more likely to be fully credit constrained (FCC) than firms elsewhere in the world (Figure 12). Fully credit constrained firms are those that have no external financing and were either rejected for a loan or did not apply because of its terms and conditions. The U-shaped relationship between firm size and the proportion of firms who are fully credit constrained supports the concept of the missing middle phenomenon in credit provision for enterprises. In Ethiopia, small-sized firms (10-20 employees) are the most credit constrained of all firm segments (57 percent), more than micro, medium, or large firms (41, 49, and 24 percent respectively).25 These results, along with the ones presented earlier on loan rates by firm size (Figure 11) support the notion that small firms struggle the most in obtaining access to finance. As the supply-side survey confirmed (Chapter 4), microfinance institutions in Ethiopia cater to microenterprises, while banks primarily have deposits by and loans to medium and large firms. In contrast, among both MFIs and banks, the shares of deposits and loans associated with small establishments are extremely small. Figure 12: Nearly half of firms in Ethiopia are fully credit constrained Figure 12: Nearly half of firms in Ethiopia are fully credit constrained 100%   90%   17   25   24%   80%   41%   46   49%   49%   46%   70%   22   57%   10%   60%   28   %  of  Firms   11%   27%   50%   23   10   5%   9%   8%   10%   9%   9%   40%   15   8   7%   13%   30%   1%   20%   38   36   38%   36%   40%   36%   36%   32   29%   10%   0%   Enterprise   Sub-­‐Saharan   Ethiopia   Micro  (0-­‐9)   Small   Medium   Large  (100+)   Age=1-­‐5   Age=6+   Surveys  110   Africa   2011   (10-­‐20)   (21-­‐99)   country   Region   average   (2006-­‐2011)   NCC   MCC   PCC   FCC   NCC=Not Note: Note: Credit NCC=Not Constrained, Credit MCC=Maybe Constrained, MCC=Maybe Credit Credit Constrained, Constrained, PCC=Partially PCC=Partially Credit Constrained, Credit Constrained, Credit FCC=FullyCredit FCC=Fully Constrained. Constrained. Source: Enterprise Surveys Source: Enterprise Surveys The relationship between firm age and credit constraint status is weaker. There is The relationship between no perceivable firm age difference and credit between constraint the degree status is weaker. of credit-constraint and There is noThis perceivable firm age. finding difference is consistent between with results the degree found using the of credit-constraint Survey and firm of Large age. Thisand Medium finding Scale is consistent with results Manufacturing found using the Surveyfrom Industries 2000-2011 of Large suggesting and Medium that Scale age is not a significant Manufacturing factor Industries from 2000- in predicting 2011 suggesting probability theage that is not aof receiving factor significant a loan for manufacturing the firms. in predicting probability of receiving a loan for manufacturing firms. Furthermore, our analysis reveals that a robust set of significantly negative performance Furthermore, characteristics our analysis area associated reveals that robust set ofwith being credit significantly constraint. negative After having performance characteristics defined firms as credit constrained if they are either partially (PCC) or fully credit are associated with being credit constraint. After having defined firms as credit constrained if they constrained (FCC) we observe that these firms tend to exhibit significantly lower rates of are either partially growth (PCC)of and levels fully credit orlabor constrained productivity (FCC) 26 we observe that these firms tend to exhibit (Table 6). A firm that is credit constrained tends to have sales growth that is 15 percentage points productivity significantly lower rates of growth and levels of labor 6).26 A firm (Table growth lower, employment is 5 is credit that that percentage points lower, and labor productivity growth that is 11 percentage points lower growth constrained tends to have sales growth that is 15 percentage points lower, employment 5 percentage that is than are not lower, firms who points Since growthgrowth and labor productivity credit constrained. that rates were is 11 percentage calculated between points fiscal lower than years firms 2009 who and are not 2011, it constrained. credit is likely that these Sincefirms’ growth decline post the rates were financialbetween crisis calculated fiscal andfurther has been years 2009 2011, exacerbated by the it is likely that lack these of credit. firms’ decline post the financial crisis has been further exacerbated by the lack of credit. 25 We also restrict the analysis to the Addis Ababa region since micro firms were only sampled in this region. When 26 do this, the trend support the missing middle phenomenon is still present. we One important caveat of these results is that these correlations should not be given a causal interpretation, 26 One important in fact caveat while it is of these possible results that credit is that these constraints correlations negatively affect should not be given firm-performance, a causal it is interpretation, also possible that in poor fact performing while firms are it is possible that creditto unable secure loans constraints because unable negatively affect to present viable projects. firm-performance, it is also possible that poor performing firms are unable to secure loans because unable to present viable projects. 44 SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 31 Table 6. Performance and Credit Constraint (1) (2 (3) (4) (5) (6) Annual Real Annual Annual Labor Log(Output Sales Growth Employment Productivity TFP 1=Job Creator Per FT Worker) Rate Growth Rate Growth Rate Firm is Credit -12.10** -3.649* -9.460* -0.681*** 0.443 -1.028 Constrained (4.665) (2.029) (4.951) (0.201) (0.620) (0.818) Made Investments 13.26*** 4.094* 6.912 0.394** 2.241*** -1.586*** (4.784) (2.222) (5.449) (0.188) (0.700) (0.594) log(Age) -3.127 -6.825*** 3.609 0.105 0.510 0.0117 (2.847) (1.542) (3.146) (0.130) (0.380) (0.0123) Exporter 0.0653 0.00192 0.0529 0.00162 -0.00517 0.00670 (0.0755) (0.0695) (0.0981) (0.00548) (0.00762) (0.00719) Female Manager 0.00697 -0.0246 0.0458 -0.00143 -0.000382 0.0216 (0.0435) (0.0205) (0.0471) (0.00179) (0.00543) (0.0180) Foreign Ownership -0.166 0.0250 -0.155 -0.00677 -0.0205** 0.0407*** (0.124) (0.0490) (0.155) (0.00820) (0.00716) (0.0102) Sole Proprietorship -0.0260 0.00488 -0.0555 -0.00315 -0.00506 0.726 (0.0593) (0.0248) (0.0703) (0.00256) (0.00486) (1.768) Constant 16.05 26.08*** -5.708 12.20*** -0.586 7.769*** -12.10** -3.649* -9.460* -0.681*** 0.443 (2.468) Observations 368 537 355 473 54 437 R-squared 0.263 0.168 0.174 0.246 0.547 Notes: Models (1)-(5) are OLS, model (6) is logit. Regional, sector, and size controls are included. Micro-sized firms are included in regression (2) and (6). Regression (5) is only manufacturing firms. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Source: Enterprise Surveys 3.3 Liquidity Constraints After having analyzed differences in terms of credit constrained using the Enterprise Survey data, this section expands the previous analysis and conducts further tests regarding firm characteristics associated with being financially constrained. This analysis relies on the Survey of Large and Medium Scale Manufacturing Industries conducted during the period 2000-2011. Following the recent literature assessing firms’ financial constraints, we evaluate the extent to which firms investments are correlated to their cash flow. To determine whether or not firms are liquidity or financially constrained, we examine if changes in investment27 and cash-flows28 are significantly correlated over time29. The basic intuition of this approach is that if a firm was unconstrained in its access to external financing, its investment patterns should not be correlated to its cash-flows. On the opposite, the existence of credit constraints would lead to a situation where the firm is only able to invest in the presence of positive cash-flows. Using this approach developed originally by the influential work of Fazzari, Hubbard and Petersen (1988) we are able to identify which types of firms are more likely to be more credit constrained. 27 Investment is calculated as the investment in fixed assets plus the depreciation 28 Cash flow = profit after taxation, extraordinary profit, and depreciation 29 For a complete discussion on the methodology, please refer to Appendix 5. 32 The literature pioneered by Fazzari et al (1988) argues that a positive investment-cash flow relationship can be interpreted as evidence of financial constraints. In a related work Fazzari et al (2000) clearly argues how constrained firms with large cost of external financing tend to have larger investments-cash flow sensitivity than relatively unconstrained firms that have very small cost of external funds. Building on this approach, a previous study by Carpenter and Petersen (2002) showed how the growth of small firms is constrained by external access to finance. Similarly, another study by Guariglia et al (2008) shows how in China while the investments of state-owned enterprises are not affected by cash flow; the opposite is true for privately-owned firms. In our analysis, we find that investment decisions of manufacturing firms in Ethiopia are heavily dependent on their cash flows (Appendix 5, Table 19). This confirms that firms are constrained in their access to external financing sources when making investments in fixed assets such as machinery or the purchase of land. In addition, we find some limited evidence that these credit constraints are stronger for young firms. However when looking at small firms only, we found that age matters with young small firms being more liquidity constrained than other small firms (Table 22). SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 33 4. SUPPLY-SIDE ANALYSIS Main research questions addressed: 1. How do banks and MFIs define MSMEs? 2. What is the actual extent of banks and MFIs’ involvement with MSME? 3. What are the main drivers and obstacles to MSME finance? 4. Are banks and MFIs’ business models adequate to support MSMEs? This section presents the results of an ad hoc survey on the supply side of MSME financing in Ethiopia addressed to commercial banks as well as microfinance institutions. The questionnaire was designed to address the following areas: (i) the extent of banks and MFIs’ involvement with MSMEs; (ii) the characteristics of the credit market for MSMEs (iii) the main drivers and obstacles to MSME finance; (iv) the adequateness of financial institutions’ business models for supporting MSMEs (including marketing strategies, specialized products and services as well as models of credit risk management). The methodology used for the supply-side study is comparable to the one used in recent studies undertaken by the World Bank to analyse SME finance in other African countries (including South Africa, Kenya, Nigeria, Rwanda, Tanzania and Benin). The survey was originally addressed to 16 financial institutions covering over 90% of the total assets in the banking and microfinance sectors. 13 financial institutions responded to the survey - 7 banks representing 87.1 percent of the banking sector asset portfolio and 6 microfinance institutions representing 70 percent of the micro finance sector asset portfolio30. Due to a varying response rates to questions the questionnaires were supplemented by structured face-to-face interviews conducted with 6 banks and 5 microfinance institutions. The results presented in this section are based on the written responses to the questionnaire as well as views expressed during interviews and discussions. Main findings from the supply-side analysis are summarized below and then discussed in-depth in the following sub-sections. i. Defining MSMEs a. Financial institutions in Ethiopia lack a commonly agreed definition of MSMEs. While the majority of MFIs uniformly uses the MSE definition that is laid out in the Government’s National MSE Development Strategy, commercial banks do not seem to uniformly distinguish among small, medium and large enterprises. ii. Financial institutions’ involvement with MSMEs a. Ethiopia is lagging behind other Sub-Saharan Africa’s countries and developing countries in terms of lending to SMEs. The share of SME lending in overall lending portfolio in Ethiopia is in fact only 7 percent, among the smallest shares in Sub- Saharan African countries as well as far below that of developing economies. b. Although MFI and bank involvement in SME lending is limited, financial institutions believe that the potential for this segment of the market is very good. The majority of surveyed financial institutions believe that prospects for the SME market are good and that the SME market size is large. The small enterprise segment is also identified as the most promising segment for growth, by both MFIs and banks. c. SMEs represent a missing middle in the financial sector: Lending to SMEs is limited 30 Originally 16 financial institutions were approached; nine banks holding 93.5 percent of the total assets of the banking sector, and seven microfinance institutions with 93 percent of total assets of micro finance sector. From this original sample, The 13 financial institutions who provided data are Commercial Bank of Ethiopia (CBE), Wegagen Bank, Dashen Bank, NIB Bank, Awash International Bank, Construction and Business Bank (CBB), Development Bank of Ethiopia (DBE), Amhara Credit and Saving Institution (ACSI), Oromia Credit and Saving Share Company (OCSSCO), Addis Credit and Saving Institution (AdSCI), Omo Micro Finance Institution, Wasasa and Wisdom MFI. SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 35 as MFI deposit and loan portfolios are comprised mainly by microenterprises (over 90 percent). The same is true for commercial bank portfolios and large enterprises. Deposits and outstanding loans to SMEs typically comprise less than 10 percent of the total portfolios of MFIs or banks. This leaves a considerable missing-middle of SMEs not served by either banks or MFIs and who need access to finance d. High heterogeneity of lending patters: MFIs issues the most number of loans to SMEs, but banks issue the most value. Seventy-four percent of SME loans outstanding values from the sample of financial institutions are by banks, only a quarter of SME loan values originate from MFIs. In terms of the number of loans, ACSI has 8,670 loans outstanding to SMEs compared to only 88 by the CBE (Dec. 2012). Banks provide large loans with longer maturity compared to much smaller sized firms lending with short maturity provided by MFIs. Interest rates for SME customers vary between 10 percent for low risk MFI customers and 15.5 percent for high risk bank customers. MSMEs face higher collateral requirements and interest rates because banks perceive them as more uncertain and harder to evaluate. As to the quality of loan portfolios, there is only a slight difference between banks and MFIs, with the latter having a non-performing loan rate of 2.3 percent and banks of 1.9 percent. iii. Main drivers to MSME finance a. All banks and MFIs indicated that expected returns and the contribution to the economic development of the country as the main drivers for lending to SMEs and microenterprises. The co-existence of the economic dimension of profitability of the business with the more political dimension of contributing to the country’s economic development represents an interesting feature of the Ethiopian market, where publicly owned financial institutions dominate both the banking and the microfinance sector. b. While banks and MFIs believe there is high growth potential in lending for small enterprises, the current lack of involvement is due to perceived risks. Most financial institutions in this study perceive costs and risks to be higher in the SME segment compared to the large enterprise segment. Banks seem to have a more negative perception of risks and costs than MFIs. Further, when asked to compare profitability of SME loans versus large enterprise loans, this is considered considerably lower in the SME segment. c. Government financed programs (credit guarantee programs and line of credit with technical assistance) are important drivers. When asked about the impact of government financed programs on the decision to engage in SME finance the picture that emerges clearly indicates that both categories (i.e. banks and MFIs) see favourably partial credit guarantee schemes and the provision of dedicated credit lines associated with technical assistance. Directed credit programs are also perceived as having a positive impact, confirming, once again, the dominant role that public institutions play in the banking and microfinance sector. d. The potentialities of the credit bureau are not effectively exploited. As shown in Figure most banks use credit bureau information for MSMEs loan analysis. 3 out of 5 banks consider the credit bureau to be effective. However only the negative information provided is used and the credit bureau information has limited input and an insignificant contribution to loan decisions. iv. Main obstacles to MSME financing a. SME-specific factors and macroeconomic factors were selected by all MFIs and banks to be significant or very significant obstacles to the development of SME lending. Regarding the SME specific factors, most of the financial institutions highlighted the poor quality of financial statements, inability to manage risk, lack of knowledge of business management, lack of awareness on how to be bankable, lack of adequate collateral and informality of SMEs as the major challenges. Regarding macroeconomic 36 aspects, inflation, exchange rate, tax regulation and high vulnerability of the agriculture sector were mentioned by financial institutions. b. The contractual environment and lack of collateral registry inhibit secured lending and constraints access to finance for SMEs. Lack of contract enforcement and judiciary inefficiency were indicated as main obstacles concerning the contractual environment. There is no legally authorized body to register machinery and/or equipment for it to be held as collateral. The absence of a collateral registry in combination with ineffective enforcement of contracts in case of default could contribute to significant losses for banks and this could have significant impact on the access to finance for SMEs. c. Changes in the market due to imposed lending restrictions are seen as a hindrance causing liquidity constraints. The regulatory framework affecting the liquidity position of banks is seen as another significant obstacle to involvement with SMEs. The majority of surveyed financial institutions reported that there have been significant changes in the market for lending to SMEs which affected banks in terms of liquidity and overall competition in the banking sector. d. MFIs are more likely than banks to perceive competition in SME lending as an obstacle. Banks are more likely to perceive bank-specific factors and characteristics of SME lending as obstacles. Bank specific factors include amongst others the lack of interest at the bank, limited geographic outreach, lack of appropriate products and knowledge on how to evaluate MSMEs or high collateral requirements. v. Adequateness of business models a. Most of surveyed financial institutions lack a dedicated and specialized MSME unit or department within their organizational structure. The organizational model used by the majority of the interviewed institutions does not take into account the need to set up a specialized MSME unit or department to better serve the MSME clientele. 9 out of the 12 financial institutions did not possess a separate SME department or unit at the time of the study. b. Loan appraisal techniques are still mostly based on traditional relationship lending rather than on transactional technologies such as credit scoring etc. When it comes to appraisal and monitoring of MSME loans, these are largely done through the establishment of a close relationship with clients for both banks and MFIs. A minority of banks stated using transactional technologies such as credit scoring risk rating tools, factoring or leasing. None of the MFIs use these techniques. c. Products are highly standardized and there is very limited product innovation. The product mix offered by financial institutions does not seem to be sufficiently large. Other lending products such as leasing and factoring are not offered by any of the financial institutions in this study. Furthermore, products provided to the SME and micro enterprise market are largely standardized and the efforts to continuously adapt them to client’s needs are limited. The great majority of financial institutions in the study reported no change in their financial product offering between the years 2010 and 2012. d. On the institutional level and regarding specific business models, MFIs and banks diverge in terms of sector targeting. Both target enterprises in the manufacturing sector as required by the government. Banks however have further outreach to the export trade sector, while MFIs focus more on agriculture. The only sector banks are not aiming to reach that MFIs are, is petty trade. e. Geographic location is not considered as an important marketing criterion and distribution channels are still mostly based on branches. Geographic location does not appear to be an important marketing criterion for financial institutions. This is particularly the case for MFIs as the 5 dominant ones, with more than 90 percent SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 37 of total asset portfolio of the micro financing sector, are affiliated with the regional governments in Ethiopia. Regarding their geographic outreach most of the banks and MFIs use only their own branches as distribution channels. f. Long term financing needs of SMEs do not seem to be properly addressed. The average loan maturity for SME loans reported by MFIs was 2.38 years while for the banks it was 6 years. According to the questionnaire responses, the average maturity loan for large enterprises was 10.4 years. This indicates that long term financing needs of SME’s are not well addressed and that there is a potential market gap here. 4.1. Defining MSME Financing in Ethiopia In order to provide an insight into how financial institutions view the MSME segment, they were asked to provide data on how they define small enterprises (SE’s) and micro enterprises (MI’s). Typically banks define SME’s according to the annual turnover of the business, loan size, number of employees and/or revenues generated by the financial institution. The definitions used by MFIs in Ethiopia are based on the Government’s micro and small enterprises definition as laid out in the national MSE development strategy (2011) (Table 7). All MFIs, besides one, uniformly use the number of employees’ criteria. Most MFIs also categorized micro and small enterprises in terms of turnover and loan size. Some of the MFIs indicated that in addition they categorized businesses according to the capital requirement as defined in the national policy document. Table 7. MSME Definitions, by National MSE Development Strategy Level of the Number of Sector Total assets enterprise Employees Micro enterprise industry <= 5 Less than or equal to 100,000 ($USD 6000 or 4500 Euro) service <= 5 Less than or equal to 50,000 ($USD 3000 or 2200 Euro) Small enterprises Industry From 6- 30 Less than or equal to 1.5 million ($USD 90,000 or 70,000 Euro) service From 6-30 Less than or equal to 500,000 ($USD 30,000 or 23,000 Euro) Source: National MSE Strategy 2011 As shown in Table 8, definitions can vary to a certain degree when using the definition by turnover and this needs to be taken into account when comparing data between microfinance institutions. The MFIs uniformly used the policy definition in the MSE Development Strategy in terms of the number of employees (1-5 employee for micro enterprise and 6-30 for small enterprises). Hence, the firms above this employment definition are categorized by default as medium and large enterprises. Medium enterprises are not clearly defined by the financial institutions. Only two MFIs defined the medium enterprises in terms of loan size which highly varies (the maximum loan size is 10 million Birr for one MFI while it is 200 thousand Birr for the other). Similarly, firms are not defined as medium enterprises and large enterprises neither by the government in terms of size of capital requirement nor by MFIs in terms of turnover and loan size. 38 Table 8. MFIs definition of Micro and SMEs, by turnover, employee size, and loan size Definition by Turnover (ETB) ACSI OCSSCO AdCSI Omo Micro Wasasa Micro < 100,000 50,000-100,000 N/A 50,000 - 75,000 Small 100,000-1.5 Million 100,000-500,000 100,000-500,000 Definition by Number of Employees ACSI OCSSCO AdCSI Omo Micro Wasasa Micro <5 <5 1 to 5 1 to 5 <5 Small 6-30 6-30 6-30 6-30 6-30 Definition by Loan Size (ETB) ACSI OCSSCO AdCSI 1,000,000 Omo Micro Wasasa 100,000 Micro < 15,000 40,000-50,000 500 - 100,000 N/A 100-10,000 Small 15,000-150,000 75,000-1 million 100001-1,000,000 10,000-100,000 In the absence of a distinction between micro, small and medium enterprises, most In the of the data absence ofcollected from a distinction the commercial between banks micro, small is in and aggregate medium form. Most enterprises, banks most in data of the this study collected thedispose do not from of data commercial records banks separatingform. is in aggregate enterprises largeMost from banks in thisSMEs. study doMost not of dispose the of MFIs data similarly records separating do not have large separate enterprises records fromof data Most SMEs. for SMEs of theand micro MFIs enterprises. similarly do not have It should separate be noted records that while of data efforts for SMEs were made and micro to ensure enterprises. completeness It should be notedof data, that data while was were efforts provided made on a voluntary to ensure basisof completeness and was data, notwas data always available provided on aas wished.basis voluntary Hence,andthe data was not in always this section available is based as wished. on what Hence, thewas datareceived and may in this section not always is based be representative on what was received and of the may not 31 whole be always banking and micro representative offinancing the whole banking.and micro financing industry31. industry 4.2. THE EXTENT OF BANKS’ AND MFIS’ INVOLVEMENT WITH 4.2. SMES The Extent of Banks’ and MFIs’ involvement with SMES section explores Thisexplores This section to what to what extent extent commercial commercial banks and MFIsand banks MFIs inare in Ethiopia Ethiopia arein the engaged engaged SME in the finance SME finance business business by looking by at their looking actual at their lending actual lending patterns. patterns. The majority of surveyed The majority financial institutions of surveyed believe that financial institutions prospects believe for the SME that prospects market for are good the SME market are good and that the SME market size is large. The small enterprise segment is most and that the SME market size is large. The small enterprise segment is also identified as the promising segment also identified for growth, as the by both MFIs most promising and banks segment (Figure 13). for growth, by both MFIs and banks (Figure 13). Figure 13. Most Promising Segments for Growth Figure 13. Most Promising Segments for Growth 6 5 5 4 4 4 3 2 2 1 1 1 1 0 Micro Small Medium Large MFI (N=6) Banks (N=7) Notes: Financial institutions were allowed to select more than one segment. Notes: Financial institutions were allowed to select more than one segment. 31 Despite repeated efforts to collect data from the selected financial institutions, Debit Credit and Saving Institutions, Bank of Abyssinia and United Bank have not provided information that could be useful additional input for the analysis in this study. 31 Despite repeated efforts to collect data from the selected financial institutions, Debit Credit and Saving Institutions, SME FINANCE INBank of Abyssinia ETHIOPIA: andTHE ADDRESSING United Bank MISSING have CHALLENGE MIDDLE not provided information that could be useful 39 additional input for the analysis in this study. However, when looking at banks and MFIs’ lending patterns, the picture that emerges from the survey is quite different, showing a very limited involvement of banks and MFIs with SMEs.32 However, when looking at banks and MFIs’ lending patterns, the picture that emerges from the quite different, survey isBefore focusing showing on SMEa very limited lending involvement patterns, of banks it is useful and MFIs to look with at the SMEs. overall 32 levels of deposits Before focusing and SME lendingloans onoutstanding of it patterns, banks andto is useful MFIs look at get tothe an idea overall levelsof of the different deposits and proportions loans outstanding existing between of banks the two and MFIs categories to get an idea ofof thefinancial different institutions. Amongbetween the 13 proportions existing financial the institutions two categories that were of financial surveyed, institutions. 12 (6the Among banks and 6 MFIs) 13 financial provided institutions data surveyed, that were on their gross 12 deposits (6 banks and 6 and outstanding MFIs) loans provided data as ofgross on their December deposits31,and2012, reported outstanding in Figure loans 14 and as of December Figure 31, 2012, 15 for MFIs reported and banks in Figure 14 and respectively. Figure 15 for TheMFIsgross and banksdeposits of the CBE, respectively. the leading The gross deposits commercial of the CBE, the bank, sixty-sevenbank, are commercial leading timesarehigher than the sixty-seven deposits times higher of the the than market leader deposits in of the microfinance, market leader in ACSI. Similarly, microfinance, outstanding ACSI. loans of CBE Similarly, outstanding loansare of more CBE arethan moreeighteen times than eighteen higher times that that than than higher of ACSI. of ACSI.Within the banking Within the sector bankingsector CBE CBE is theis unquestioned the unquestioned marketmarket leader. This indicates that the banking sector is dominating both deposit mobilization leader. This indicates that the banking sector is dominating both deposit mobilization and and the lending industry the lendingin Ethiopia. industry in Ethiopia. Figure 14. MFIs Figure Gross 14. MFIs Deposits Gross and Deposits and Outstanding Outstanding LoansLoans in in Billion Billion Birr Birr (Dec. 31, (Dec. 31, 2012) 2012) 4.0 3.3 3.0 2.0 1.7 2.0 1.0 1.1 0.8 1.0 0.4 0.3 0.2 0.0 0.2 0.1 0.0 ACSI OCSSCO Adds* Omo Micro Wasasa Wisdom MFI Total loans outstanding (Dec 2012) Total deposits outstanding (Dec 2012) Figure 15. Banks Figure Gross 15. Banks Deposits Gross and Deposits and Outstanding Outstanding in Billionin Loans Loans Billion Birr Birr (Dec. 31, (Dec. 31, 2012) 2012) 150.0 134.4 100.0 61.7 50.0 4.3 6.3 9.1 14.7 4.5 6.3 6.2 10.3 1.9 3.9 0.0 CBE Wegagen Dashen Bank NIB Awash CBB DBE Bank 32 While all microfinance institutions state having SMEs and micro enterprises as clients, half of surveyed to loans Total banks declare not serve outstanding (Dec SMEs or micro 2012) enterprises Total deposits outstanding (Dec 2012) at all. Among these levels of deposits and loans, the space dedicated to SMEs in the financial sector is Among quite limited. Thethese levels share of of deposits SME lending in theand 54loans, overall the space lending dedicated portfolio to is in Ethiopia SMEs in factin the only 7 financial percent, sectorthe among is quite limited. smallest The shares share of SME in Sub-Saharan lending African in the as countries overall lending well as far belowportfolio that of in Ethiopia developing is in fact (Figure economies percent, only 716). share the among Nigeria’s smallest of SME shares lending lies at in Sub-Saharan 5 percent, African South Africa’s at countries 8 aspercent percent, 16 far below well asacross that of developing developing economies, andeconomies 22 percent(Figure 16). Nigeria’s in developed share countries (Fuchs ofalSME et et al lies lending 2011; Beck 2008).at 33 5 percent, South Africa’s at 8 percent, 16 percent across developing economies, and 22 percent in developed countries (Fuchs et al 2011; Beck et al 2008).33 32 While all microfinance institutions state having SMEs and micro enterprises as clients, half of surveyed banks declare not to Figure serve SMEs16. Share or micro of SMEs enterprises at all.Lending in Overall Portfolio 33 The developing country average is based on analysis of data on 45 countries included in the Beck et al (2008) study. 25% 22% 20% 40 16% 15% Among these levels of deposits and loans, the space dedicated to SMEs in the financial sector is quite limited. The share of SME lending in the overall lending portfolio in Ethiopia is in fact only 7 percent, among the smallest shares in Sub-Saharan African countries as well as far below that of developing economies (Figure 16). Nigeria’s share of SME lending lies at 5 percent, South Africa’s at 8 percent, 16 percent across developing economies, and 22 percent in developed countries (Fuchs et al 2011; Beck et al 2008).33 Figure 16. Share of SMEs Lending in Overall Portfolio Figure 16. Share of SMEs Lending in Overall Portfolio 25% 22% 20% 16% 15% 10% 8% 7% 5% 5% 0% Nigeria Ethiopia South Africa Developing Developed Countries Countries Source: Analysis Lendingand of survey responses SME lending. SMEs Berg to reports; et al (2012), is limited Fuchstarget as MFIs et al (2011), Beck et al (2008) microenterprises and 34 et al (2008) Source: Analysis of survey responses and reports; Berg et al (2012), Fuchs et al (2011), Beck bank clientele are primarily large enterprises (Figure 17). The 5 MFIs who reported 35 lending SME figures lending. disaggregated Lending to SMEs isby size clientas limited focus MFIs target their lending on microenterprises microenterprises and bank clientele ; 92 are percent large primarily of their total loans enterprises are (Figure disbursed 17). The 5 MFIs to who microenterprises 34 reported lendingwhile only figures 8 percent are disaggregated by client size focus their lending on microenterprises issued to SMEs. Among banks, only the CBE reported disaggregated lending by client 35 ; 92 percent of their total loans are disbursed to size. microenterprises The CBE tends while percent only 8 on to focus largeare issued to SMEs. enterprises Among banks, and provides lending to the only the CBE SME reported sector 33 disaggregated   The lending developing country by client average size. is The based on CBE tends analysis comprising almost 6 percent of the bank’s total disbursements. of to datafocus on on 45 large countriesenterprises included in and the provides Beck et al (2008) study. lending to the SME sector comprising almost 6 percent of the bank’s total disbursements. 55 Figure 17. The Missing Middle: Lending to SMEs is limited Figure 17. The Missing Middle: Lending to SMEs is limited 92.0% 94.32% 100.0% Proportion of Total Lending 80.0% 60.0% 40.0% 20.0% 8.0% 5.68% 0.0% Microenterprise SME Large MFIs (N=5) CBE By zooming on the reduced share of lending that both MFIs and banks provide to SMEs, substantial disproportions emerge when comparing actual values with actual numbers of outstanding loans. By zooming on the reduced share of lending that both MFIs and banks provide to In terms of the value of loans, the CBE also overshadows MFIs in the total value of disbursements. SMEs, substantial While representing disproportions only 5.7 percent ofemerge when the CBE’s comparing total lending actual portfolio, values the with of actual value actual SME numbers lending of outstanding totaled Birr as In 1.6 billion loans. terms of end of the 2012 value (Table of loans, 9). The amounttheofCBE SME also overshadows lending for all MFIs MFIs in the combined total totaled value only of disbursements. 231 million Birr. While representing only 5.7 percent of the CBE’s total lending portfolio, the actual value of SME lending totaled 1.6 billion Birr as of end 2012 (Table 9). The amount of SME lending for all MFIs combined totaled only 231 million Birr. Table 9. Total Lending (Dec. 2012, Birr) MFIs (N=5) CBE Micro ETB 2,667,957,137 34 The five MFIs are: OCSSCO, Adds*, Omo Micro, Wasasa, and Wisdom SME ETB 231,005,054 ETB 1,577,708,662 35 Most MFIs define microenterprises to be those with less than five employees. Large ETB 38,359,274,933 SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE The imbalances in the provision of loans to SMEs vis-a-vis micro and large41 enterprises is further highlighted by the distribution in the actual number of loans Table 9. Total Lending (Dec. 2012, Birr) MFIs (N=5) CBE Micro ETB 2,667,957,137 SME ETB 231,005,054 ETB 1,577,708,662 Large ETB 38,359,274,933 question, it is clear that the bulk of MFI loans are issued to microenterprises, and The imbalances virtually in the all loans provision from of loans banks are toto issued SMEs vis-a-vis large micro and large enterprises is further companies. highlighted by the distribution in the actual number of loans outstanding by client size (Table 10). Among financial institutions who responded to this question, it is clear that the bulk of MFI loans are issued to microenterprises, and virtually all loans from banks are issued to large companies. Table 10. Number of Loans Outstanding (Dec. 31, 2012) MFI(Dec. 31, 2012) Table 10. Number of Loans Outstanding Banks ACSI Wasasa CBE Wegagen Dashen Bank Micro 781,954 MFI 63,280 Banks SME 8,670 ACSI 262 Wasasa 88 CBE Wegagen Dashen Bank Large Micro 781,954 63,280 77,245 4,248 7,284 Total 790,624 63,542 77,333 4,248 7,284 SME 8,670 262 88 Large 77,245 4,248 7,284 790,624 of the number Total The distribution 63,542 of loans and loan value 77,333 to SMEs also 4,248 highlights an 7,284 important aspect that should be noted for this segment. The CBE issued $1.6 billion Birr The distribution in loans to 88 of while of the number SMEs, and loan loansissued MFIs value only to SMEs $231 also Birr million highlights an important to almost aspect 9,000 SMEs. that should be noted for this segment. The CBE issued $1.6 billion Birr in loans Clearly, there is great heterogeneity in the population of SMEs, some who are receiving to 88 SMEs, while MFIs issued only $231 million Birr to almost 9,000 SMEs. Clearly, there is great very large loans and are perhaps well-established, and others who are receiving small heterogeneity in the population of SMEs, some who are receiving very large loans and are perhaps well-established, loans. and others who are receiving small loans. SME SME SME deposits. deposits. deposit SME deposit is mobilization mobilization also dominatedis also dominated by the by the banking sector banking while sector MFIs provide whilecredit more MFIswith provide more smaller loan size. with credit smaller The level loan size. of deposits The from level SMEs inof thedeposits selectedfrom MFIsSMEs (ACSI, in the selected OCSSCO, MFIs and Wasasa) (ACSI, range from OCSSCO, and Wasasa) 7 to 15 percent while only range fromfor 1 percent to 15 7 CBE percent 18).while (Figure While 1 percent onlydeposits SME for CBE comprise (Figure only 18).of 1 percent While all CBESME deposits deposits, comprise the total volume only 1 percent of CBE of all SME deposits CBE deposits, equals 86 percent the total of all SME volume deposits. CBE ofACSI SME holds thedeposits equalsamount second largest 86 percent of SMEofdeposits all SME (8 deposits. ACSI holds the second largest amount of SME deposits (8 percent). This percent). This confirms the finding that CBE primarily serves large firms, while micro enterprises confirms utilize MFIs.the finding It also that the underlines CBE primarily potential serves large for cross-selling firms, while of lending productsmicro enterprises to SME savers for both CBE and ACSI. utilize MFIs. It also underlines the potential for cross-selling of lending products to SME savers for both CBE and ACSI. Figure 18: Proportion of Outstanding Deposits, by Client Size Figure 18: Proportion of Outstanding Deposits, by Client Size 150 Proportion of Total 99 89 Deposits (%) 100 50 11 1 0 Microenterprise SME Large MFIs (N=3) CBE Loans terms and conditions. Figure 19 shows a comparison of average interest rates applied to loans by banks and MFIs. Banks best rate for SME’s was reported at 11.20 percent and almost equal Loans to the best MFIs and terms rate of conditions at 11.50 Figure percent..The shows 19 rates highest a SME’s for comparison of however average diverge interest with banks rates applied to loans by banks and MFIs. Banks charging up to 14.50 percent and MFIs only 11.80 percent.best rate for SME’s was reported at 57 42 11.20 percent and almost equal to the best rate of MFIs at 11.50 percent. The highest rates for SME’s however diverge with banks charging up to 14.50 percent and MFIs only 11.80 percent. Figure 19. Average Interest Rates as at December 2012 Figure 19. Average Interest Rates as at December 2012 20 14 14.5 15 11.5 11.8 11.2 9.5 10 5 0 Group Rate SME Best Rate SME Highest Rate Microfinance institutions Banks Notes: Group Rate for MI are for MSMEs, group rate for Banks are for large enterprises. Notes: Group Rate for MI are for MSMEs, group rate for Banks are for large enterprises. Banks as well Banks as microfinance institutions reported as well as microfinance providing institutions all SME reported loans in all providing local currency. loansThe SME in average loan maturity for SME loans reported by MFIs was 2.38 years while local currency. The average loan maturity for SME loans reported by MFIs was for the banks it was 2.386 years. According to the questionnaire responses, the average maturity loan for large years while for the banks it was 6 years. According to the questionnaire responses, the enterprises was 10.4 years. average This indicates maturity loan for that largelong term financing enterprises needs was 10.4 of SME’s years. Thisare not wellthat indicates addressed and long term that there is a potential market gap here. financing needs of SME’s are not well addressed and that there is a potential market gap here. According to the responses provided, non-performing loans continue to be well managed by interviewed financial institutions. The majority of financial institutions (5 out of 6 MFIs, and 4 out of 5 banks)According maintained totheir responses provided, the non-performing non-performing loan ratio below 5 percent loans continue which to beset is the target well by managed by interviewed financial institutions. The majority of financial institutions the NBE. One bank reported being at 5 percent and one MFI slightly above the 5 percent target. For (5 out of most ofthe banks and 6 MFIs, one 4 out day was 5 banks) maintained of considered the threshold their non-performing after which SME loans loan wereratio below 5 categorized percentNevertheless, overdue. which is theonlytarget setof 1 out by the NBE. 4 banks One in this bank study havereported being a dedicated at 5 loan percent recovery andThe one unit. 3 MFI MFIs slightly who above answered thisthe 5 percent question target. reported thatFor they most of the did have banks a loan one day recovery unit was considered or it was handled bythe threshold their after operations which SME department. The loans were categorized percentage of recovery to overdue. Nevertheless, loan value for SMEs ranges only 1 out from 20 toof 4 percent. 100 banks inOne study thisbank have a a reported dedicated recovery costloanestimated recovery at unit. 7.41The 3 MFIs who answered percent. this question reported that they did have a loan recovery unit or it was handled by their operations department. The percentage of recovery to loan value for SMEs ranges from 20 to 100 percent. One bank reported a recovery cost estimated at 7.41 percent. 4.3. Drivers and obstacles to SME Financing The previous section highlighted on one hand the positive perception that banks and MFIs have with 4.3.reference to small enterprises DRIVERS as the most promising AND OBSTACLES TO SME segment for growth and on the other FINANCING hand the existence of the missing middle phenomenon in terms of SME lending volumes. Against this background, the current The previous section section investigates highlighted on oneare what the the hand main drivers that positive would trigger perception banks that banks and MFIs’ interest in engaging in SME lending and also what are the biggest and MFIs have with reference to small enterprises as the most promising segment forperceived obstacles that are preventing growth and on the them to engage other hand the fully in this market existence of the segment. missing middle phenomenon in terms of SME lending volumes. Against this background, the current section investigates what the main are 4.3.1. drivers Drivers would trigger banks and MFIs’ interest in engaging in SME that Finance of SME lending and also what are the biggest perceived obstacles that are preventing them to Main drivers. Banks and MFIs were asked to rank a set of potential drivers as either not significant, engage fully in this market segment. significant, or very significant, and to provide a rationale for their responses (Figure 20). All banks and 58 the contribution to the economic development MFIs (a total of 8) indicated that expected returns and of the country as the main drivers for lending to SMEs and microenterprises. The co-existence of the economic dimension of business profitability with the more political dimension of contributing to the country’s economic development represents an interesting feature of the Ethiopian market. Publicly owned financial institutions dominate both the banking and the microfinance sector in Ethiopia. This also explains why competition is not seen as a major driver for involvement with MSMEs given the high concentration of the market in favor of publicly owned financial institutions. Structural changes in the market were mostly seen as a driver by MFIs due to recent government’s efforts in setting-up new institutions in underserved regions of Ethiopia. SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 43 economic development represents an interesting feature of the Ethiopian market. Publicly owned financial institutions dominate both the banking and the microfinance sector in Ethiopia. This also explains why competition is not seen as a major driver for involvement with MSMEs given the high concentration of the market in favor of publicly owned financial institutions. Structural changes in the market were mostly seen as a driver by MFIs due to recent government’s efforts in setting-up new institutions in underserved regions of Ethiopia. Figure 20. Figure 20.The Main The Main Drivers Drivers of banks’ of banks’ and and MFIs’ MFIs’ with involvement involvement with SMEs and Micro SMEs and Enterprises Micro Enterprises Availability of credit lines Contributing of economic development Growth prospects in the segment over the next Structural changes in market Cross selling Supply chain links Excessive exposure to retail customers Excessive exposure to large enterprises Competition for retail customers Competition for large enterprises Returns 0 1 2 3 4 5 6 7 8 9 Not Significant Significant Very Significant Cost, profitability, and risk of SME lending. Banks and MFIs were also asked to provide their Cost, comparative profitability, assessment and of the risk cost, of SME and profitability lending. Banks risk of and MFIs SME lending were alsoto as compared asked large to provide their comparative assessment of the cost, profitability and risk of SME lending enterprise lending (Figure 21). Most financial institutions in this study perceive costs and risks to be as compared higher to segment in the SME large enterprise compared lending (Figure to the large 21). Most enterprise financial segment. Banks institutions seem to have in this a more study perceive costs and risks to be higher in the SME segment compared to negative perception of risks and costs than MFIs. Further, when asked to compare profitability of the large enterprise SME large Banks segment. loans versus seem enterprise to have loans, this is a more negative considered perception considerably lower of risks in the SME and costs segment. than MFIs. Further, when asked to compare profitability of SME loans versus large enterprise loans, this is considered considerably lower in the SME segment. Figure 21. Banks & MFIs: Comparison of Cost, Risk and Profitability of SME Loans versus Large Enterprise Figure 21. Banks & MFIs: Comparison 59 of Cost, Risk and Profitability of SME Loans Loans versus Large Enterprise Loans 4 More 4 Banks (N=6) 1 Equally 1 1 Less 2 5 3 More 3 MFIs (N=5) Equally 1 2 Less 2 4 0 1 2 3 4 5 6 Cost Risk Profitability The role of regulatory requirements. The role played by current prudential regulations on driving MSME finance decision was also tested. Responses indicate that for the great majority of institutions, prudential regulations have a positive impact on the decision to engage in SME finance. The majority of MFIs and banks rated the burden posed by regulatory documentation requirements as appropriate and beneficial (Figure 22) with only 2 banks perceiving that the regulatory documentation requirements imposed by the central bank are excessive and too stringent for SMEs. 44 2 Less 2 4 0 1 2 3 4 5 6 Cost Risk Profitability The role of regulatory requirements. The role played by current prudential regulations on driving MSME finance decision was also tested. Responses indicate that for the The role great majority of of regulatory institutions, prudential requirements. regulations The role played have by current a positiveregulations prudential the impact on on decision driving to engage MSME financein was also The SME finance. decision majority tested. Responsesof MFIs and indicate banks that rated for the the great burden majority posed of by regulatory institutions, documentation prudential requirements regulations have as appropriate a positive impact and beneficial to engage (Figure on the decision in SME 22) with finance. Theonly 2 banks majority of MFIsperceiving theburden that the and banks rated poseddocumentation regulatory requirements by regulatory documentation imposed by the requirements central bank as appropriate arebeneficial and and too excessive(Figure stringent 22) for2SMEs. with only banks perceiving that the regulatory documentation requirements imposed by the central bank are excessive and too stringent for SMEs. Figure 22. Burden posed by regulatory documentation requirements for lending to Figure 22. Burden posed by regulatory documentation requirements for lending to MSMEs MSMEs 5 4 4 3 3 2 2 1 0 - - Appropriate and No Effect Excessive almost beneficial for all products MFIs (N=4) Banks (N=5) The The role of role of Government Government financed financed programs. When programs. asked about When asked the impact about the impact of government financedof government programs financed on the decision programs to engage on SMEdecision in the finance itto engage emerges in SME clearly finance that both it emerges categories (i.e. clearly banks andthat both MFIs) have categories a positive (i.e. banks of perception and MFIs) partial have credit a positive guarantee perception schemes and the of partial provision credit of credit schemes guarantee dedicated and the lines associated with technical of provision dedicated assistance. creditcredit Directed associated linesprograms are with also technical as perceived assistance. Directed having a positive programs creditconfirming, impact, 60 areonce also perceived again, as having the dominant a positive role that public impact, confirming, institutions once again, play in the banking the dominant and microfinance role that public institutions play in the sector. banking and microfinance sector. Figure 23. Impact of government programs on SME finance Figure 23. Impact of government programs on SME finance regulatory subsidies Banks (N=3 to 6) credit line and technical assistance direct credit programmes guarantees interest subsidies Positive Neutral regulatory subsidies Negative MFIs (N=3 to 6) credit line and technical assistance direct credit programmes guarantees interest subsidies 0 1 2 3 4 5 6 7 The role of the credit information bureau. As shown in Figure 24 most banks use credit bureau information for MSMEs loan analysis. 3 out of 5 banks consider the credit bureau to be effective. However only the negative information provided is used and the credit bureau information has limited input and an insignificant contribution to loan decisions. SME Value ADDRESSING FINANCE IN ETHIOPIA: added services such THE MISSING as CHALLENGE MIDDLE credit scoring are not available at the credit 45 bureau. Currently, banks are submitting their new credit information to the credit bureau on a monthly basis. The credit bureau validates this credit information and sends it back The role of the credit information bureau. As shown in Figure 24 most banks use credit bureau information for MSMEs loan analysis. 3 out of 5 banks consider the credit bureau to be effective. However only the negative information provided is used and the credit bureau information has limited input and an insignificant contribution to loan decisions. Value added services such as credit scoring are not available at the credit bureau. Currently, banks are submitting their new credit information to the credit bureau on a monthly basis. The credit bureau validates this credit information and sends it back to the banks. Each individual bank can only access its own borrowers’ credit information and the central, institution wide data is only available at the credit bureau level. Although the technology infrastructure of the credit information system is capable of accommodating the credit information of MFIs and MFIs have been involved in the setting-up of the credit bureau, they are not using the credit information system. This is due to a lack of technology platforms in their institutions not allowing them to connect to the credit bureau. Figure 24. Importance of credit bureaus for Banks’ MSME finance Figure 24. Importance of credit bureaus for Banks’ MSME finance Are credit bureaus effective? No Yes Use of credit bureaus for MSME loan analysis 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 4.3.2. Obstacles to SME Financing 4.3.2. The OBSTACLES survey TO SME also investigated theF INANCING main obstacles to SME financing by asking financial institutions to rank a set of potential obstacles to SME finance as either not significant, significant, or very The significant, and survey also to provide investigated a rationale the for their main obstacles responses. to SME Additionally, financing a central by in-depth focus of the asking financial institutions discussions to rank with interviewed of potential a set institutions financial obstacles evolved to SME around financeto the obstacles as either SME not finance. significant, The number of significant, respondents very answered or who significant, and that to provide these a rationale factors were foror significant their very responses. significant Additionally, obstacles a central is illustrated focus in Figure 25.of the in-depth discussions with interviewed financial institutions evolved around the obstacles to SME finance. The number of respondents SME-specific who answered factors macroeconomic andthese that factors. factors were SME-specific significant very and or factors macroeconomic significant factors obstacles is were selected by all MFIs illustrated in Figure 25. and banks as significant or very significant obstacles to the development of SME lending. Regarding the SME specific factors, most of the financial institutions highlighted quality of financial the poorSME-specific statements, factors inability to manage and macroeconomic risk, lack factors. of knowledge SME-specific of business factors and management, lack of awareness on how to be bankable, lack of adequate collateral and informality macroeconomic factors were selected by all MFIs and banks as significant or very of SMEs as the major challenges. Regarding macroeconomic aspects, inflation, tax regulation and significant obstacles to the development of SME lending. Regarding the SME specific high vulnerability of the agriculture sector were mentioned by financial institutions. factors, most of the financial institutions highlighted the poor quality of financial statements, The inability contractual to manage environment andrisk, lack the lack of of knowledge a collateral of business registry. Lack ofmanagement, lack of contract enforcement awareness and judiciaryoninefficiency were also lack how to be bankable, of adequate indicated as main collateral obstacles and informality concerning theof SMEs as contractual the major challenges. environment. Regarding There is no legally macroeconomic authorized aspects, body to register inflation, machinery and/ortax regulation equipment forand it to high vulnerability of the agriculture sector were mentioned by financial institutions. be held as collateral. Therefore, issues relating to collateral and weak contract enforcement inhibit secured lending and constraints access to finance for SMEs by posing high risks to the lenders. The absence The contractual of a collateral environment registry in combination and theineffective with collateral registry lack of aenforcement . Lack of contracts of in case contract enforcement and judiciary inefficiency were also indicated as main obstacles of default could contribute to significant losses for banks and this could have significant impact on access to finance concerning for SMEs. environment. There is no legally authorized body to register the contractual machinery and/or equipment for it to be held as collateral. Therefore, issues relating to collateral and weak contract enforcement inhibit secured lending and constraints access to finance for SMEs by posing high risks to the lenders. The absence of a collateral registry in combination with ineffective enforcement of contracts in case of default could contribute to significant losses for banks and this could have significant impact on access to finance for SMEs. 46 Figure 25. Obstacles to Banks & MFIs Involvement with SMEs Figure 25. Obstacles to Banks & MFIs Involvement with SMEs (Significant and Very Significant) (Significant and Very Significant) 2 lack of adequate demand 0 2 competition in the SME market 4 4 characteristics of SME lending 2 4 SME specific factors 4 4 bank specific factors 1 4 contractual environment 3 3 legal framework affecting financial institutions 3 4 macroeconomic factors 4 0 1 2 3 4 Banks (N=4) MFIs (N=4) The legal and regulatory framework. Another significant obstacle to involvement with SMEs indicated was the legal and regulatory framework affecting financial institutions: 6 out of The legal and regulatory framework. Another significant obstacle to 10 financial institutions reported that there have been significant changes in the market for involvement with SMEs indicated was the legal and regulatory framework affecting lending to SMEs which affected banks in terms of liquidity and overall competition in the financial institutions: 6 out of 10 financial institutions reported that there have been banking sector. Banks and MFIs reported facing weak liquidity positions due to credit limits significant changes in the market for lending to SMEs which affected banks in terms of for SME and micro enterprise loans, not being able to go beyond 1 percent of their capital for liquidity and overall competition in the banking sector. Banks and MFIs reported facing microfinances institutions and 25 percent for banks. Financial institutions are required to set weak liquidity positions due to credit limits for SME and micro enterprise loans, not their lending portfolio for monitoring purpose by the NBE. These lending restrictions were being able to go beyond 1 percent of their capital for microfinances institutions and 25 imposed on private banks and then replaced by an NBE directive requiring commercial private percent for banks. Financial institutions are required to set their lending portfolio for banks to allocate 27 percent of their loan disbursements to purchase fixed and low interest monitoring purpose by the NBE. These lending restrictions were imposed on private bearing NBE Bills. banks and then replaced by an NBE directive requiring commercial private banks to allocate 27 According topercent privateof their loan disbursements commercial to purchase banks, this directive has hadfixed and low impact a negative interestonbearing their NBE Bills. liquidity and lending capacity and they are therefore not able to lend as much as they want. In a constrained liquidity environment banks are likely to favor existing, established clients when allocatingAccording to private loans as opposed tocommercial newer, riskier banks, SMEs.this Also,directive has had larger clients a better offer negative impact prospects their on fee for liquidity income. and lending Although capacity a temporary and they solution was are therefore provided by not able to lend the regulator, as much National as Bank they of want.by Ethiopia, In a constrained reducing liquidity the reserve environment and liquidity banks on requirements are likely to banks, commercial favor lowering existing, established the clients when reserve requirement allocating down from 10 loans to 5 as opposed percent andto newer, the riskier liquidity SMEs. Also, fromlarger requirement 25 to clients 20 offer percent, thebetter prospects liquidity problemforoffee theincome. private Although a temporary banks appears to still besolution an issue.was Theprovided recently by the regulator, launched government National housing Bank Ethiopia, of adds project to thisby reducing liquidity the reserve challenge and banks for private liquidity as their customers tend to withdraw their funds to deposit them in public commercial banks for long-term housing plans. 63 Competition. The obstacle given the lowest importance by interviewed financial institutions was competition in SMEs market which Ethiopia is lacking due to high concentration of few financial institutions. Furthermore, the NBE Directive No. SBB/50/2011 which raised the minimum paid up capital for establishing a bank from Birr 75 million to Birr 500 million has discouraged new entrants and forced existing ones to dissolve, thereby affecting the competitiveness of the banking sector. SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 47 Competition. The obstacle given the lowest importance by interviewed financial institutions was competition in SMEs market which Ethiopia is lacking due to high concentration of few financial institutions. Furthermore, the NBE Directive No. SBB/50/2011 which raised the minimum paid up capital for establishing a bank from Birr 75 million to Birr 500 million has discouraged new entrants and forced existing ones to dissolve, thereby affecting the competitiveness of the banking sector. 4.4. THE BANKS’ AND MFIS’ BUSINESS MODELS 4.4. The Banks’ and MFIs’ Business Models After After having having the assessed actual the assessed actual extent extent of banks’ of MFIs’ and banks’ and MFIs’in involvement involvement SME financinginand SME its financing and its main drivers and obstacles, the current section looks at the existing main drivers and obstacles, the current section looks at the existing business models of banks and business MFIs models and their of banks to and adequateness MFIs and effectively their address adequateness the to effectively SME market. The address questionnaire the included a SME market. The questionnaire included a range of questions regarding organizational range of questions regarding organizational models, product marketing, credit risk management, models, and product bad loan marketing, credit risk management, and bad loan recovery. recovery. 4.4.1. ORGANIZATIONAL MODELS 4.4.1. Organizational Models Separate Separate MSME unit. MSME unit. The model The organizational organizational used by themodel used majority byinterviewed of the the majority of the institutions interviewed institutions does not seem to take into account the need for a specialized does not seem to take into account the need for a specialized MSME unit or department to better MSME serve theunit MSME department or clientele. to better 9 out serve of the 12 the MSME financial clientele. institutions 9 out did not of the possess 12 financial a separate SME institutions or department did at possess not unit the time a ofseparate the studySME (Figuredepartment 26). or unit at the time of the study (Figure 26). Figure 26. Figure 26.Having a separate Having a separate MSME MSME department/unit department/unit to manage to manage SMEs and MicroSMEs and Enterprise Micro clients Enterprise clients No MFIs (n=6) Banks (n=6) Yes 0 1 2 3 4 5 6 Although most MFIs Although state most involved beingstate MFIs with being SMEs, only involved with2 SMEs, their2 indicateonly client relationships indicate are their client managed through a dedicated MSME unit. This is the case for only 1 Bank. relationships are managed through a dedicated MSME unit. This is the case for only 1 Bank. risk management. Most of the surveyed institutions do not separate the credit risk Credit Credit risk management. Most of the surveyed institutions do not separate the management function from credit risk management the sales function function. from In only the sales a minority function. In of respondents only (i.e. of one bank a minority and two MFIs), risk management is done 64 primarily at the branch level. Mostly respondents (i.e. one bank and two MFIs), risk management is done primarily at the risk management in banks is handled by a credit analyst and the assessment processes are not automated. branch level. Mostly risk management in banks is handled by a credit analyst and the assessment processes are not automated. Figure 27. Organization of credit risk function (Banks and MFIs) Figure 27. Organization of credit risk function (Banks and MFIs) 4 3 3 3 2 2 2 2 2 1 1 1 1 0 0 0 0 Separated from Combined Largely Mostly done by Done primarily Done primarily sales sales & risk automated credit analyst at HQ at branch level Banks (out of 7) MFIs (out of 6) Loan appraisal and monitoring. When it comes to appraisal and monitoring of MSME loans, these are largely done through the establishment of a close relationship with clients for both banks and MFIs. This is the case for 4 out of 5 banks and all MFIs who maintain close contact with their clients via on-site visits, and have continuous interaction and frequent reporting requirements. 3 MFIs and 2 banks also reported using relationship lending loan approval techniques i.e. based on soft information gathered by 48 loan officers via direct personal contact with the MSME, the owner, manger and community it operates in. 2 banks also stated using transactional technologies such as 1 1 1 1 0 0 0 0 Separated from Combined Largely Mostly done by Done primarily Done primarily sales sales & risk automated credit analyst at HQ at branch level Banks (out of 7) MFIs (out of 6) Loan appraisal and monitoring. When it comes to appraisal and monitoring of MSME loans, these Loan appraisal and monitoring. When it comes to appraisal and monitoring of largely done are MSME loans,through these are establishment thelargely of a close done through therelationship with establishment ofclients forrelationship a close both banks and MFIs. This is the case for 4 out of 5 banks and all MFIs who maintain close with clients for both banks and MFIs. This is the case for 4 out of 5 banks and all contact with their MFIsclients on-site via who visits, and have maintain close contact with their clients via on-site visits, and have continuousMFIs continuous interaction and frequent reporting requirements. 3 and 2 banks interaction also and frequentusing reported relationship reporting lending requirements. 3 loan MFIs approval techniques and 2 banks i.e. based also reported on soft using information gathered by loan officers via direct personal contact with the MSME, relationship lending loan approval techniques i.e. based on soft information gathered by the owner, manger and officersit via community loan directin. operates 2 banks also personal stated contact using with MSME, thetechnologies thetransactional owner, manger such as andcredit scoring, risk rating tools, factoring or leasing. None of the MFIs use this technique. community it operates in. 2 banks also stated using transactional technologies such as credit scoring, risk rating tools, factoring or leasing. None of the MFIs use this technique. Figure 28. Loan origination and monitoring of micro-enterprise and SME loans Figure 28. Loan origination and monitoring of micro-enterprise and SME loans 5 4 4 4 3 3 2 2 2 1 0 0 Approves through relationship Approves through transactional Close relation with client to lending technologies monitor loans MFIs (N=4) Banks (N=5) Table 11 below summarized the main features of the loan appraisal process and lending technologies reported by surveyed institutions. 65 Table 11. Business models of banks and MFIs MFIs Banks Loan appraisal process and lending technologies Information used in Combination of hard and soft information Combination of hard and soft information credit appraisal on the business and entrepreneur and information on credit history from own financial institution Qualitative assessment of SWOT analysis, Similar to MFIs qualitative and quantitative viability of the business idea, realistic assessment. In addition, banks rate the assessment of market size and potential and quality of SME management entrepreneurial character of the owner. Quantitative assessment including financial Rely on financial assessment of the analysis of the business and the sector trend. businesses and character of the owner. Lending technology Relationship based lending Relationship based lending No scoring models in place No scoring models in place Banks and MFIs equally use both quantitative and qualitative assessments for their credit analysis. All criteria of qualitative assessments including rating the quality of the SMEs management, doing a SWOT analysis, evaluating the viability of the business idea, assessing the target market size and potential as well as the entrepreneurship skills of the owner are used. The majority of banks and MFIs use qualitative techniques for credit assessment (Figure 29). The viability of the business idea and the entrepreneurial character of the owner are the two most important variables used by banks as well as MFIs. MFIs however, rely less on SWOT analysis and more on the evaluation of the market size and potential of a business than banks. SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 49 Figure 29. Use of qualitative assessment and variables for credit analysis of micro-enterprise and Figure 29. Use of qualitative assessment and variables for credit analysis of micro- SME loans enterprise and SME loans Figure 29. Use of qualitative assessment and variables for credit analysis of micro- 4 4 4 4 enterprise and SME loans 4 4 4 4 4 3 3 4 4 3 3 4 4 4 3 3 3 3 3 2 3 2 2 2 1 0 1 0 Credit analysis Rating0 the quality SWOT analysis Viability of the Realistic Entrepreneurial 0 relies on of MSME of MSME business idea assessment of character of the Credit analysis Rating qualitative Realistic the quality SWOT analysis Viability of the market management size and Entrepreneurial owner relies on assessments of MSME of MSME business idea potential of assessment character of the qualitative management MFIs (N=4) Banks (N=4) market size and owner assessments potential MFIs (N=4) Banks (N=4) MFIs and Banks MFIs both and commonly Banks utilize quantitative both commonly measures tomeasures utilize quantitative assess credit (Figurecredit 30). to assess (Figure 30). assessments Quantitative Quantitative including assessments financial analysis including of SMEs, financial projected analysis sector of SMEs, trends projected and MFIs and indicators Banks andindicators financialboth commonly analysis of SME utilize owners quantitative are equally measures used to assess forequally credit credit analysis. sector trends and and financial analysis of SME owners are used for (Figure The most 30). Quantitative important assessments criteria for SME including loan decision financial making analysis for banks of SMEs, and MFIs projected are the credit analysis. The most important criteria for SME loan decision making for banks and sector trends SME’s and indicators ownfinancial and analysis of SME ownersof are equally used for MFIs credit are the history SME’s with the credit historyinstitution, with the own financial assessment institution, financial the business assessment and of the credit the analysis. The most important criteria for SME loan decision making for banks and business and the characteristics characteristics of the SMEs owners. of the SMEs owners. MFIs are the SME’s credit history with the own institution, financial assessment of the business Figure and 30. theof Use quantitative of characteristics the SMEsand assessment owners. variables for credit analysis of micro- Figure 30. Use of quantitative assessment and variables enterprise and SME credit analysis of micro-enterprise and for loans Figure 30. Use of quantitative assessment and SME loans variables for credit analysis of micro- 6 enterprise and SME loans 5 5 6 4 5 4 4 5 4 3 4 3 3 3 3 3 4 3 3 3 3 2 3 2 3 1 2 1 2 1 0 1 Credit analysis relies Financial analysis of Projected sector Financial projection Financial analysis of 0 on quantitative MSME trends/indicators of MSME MSME owner Credit analysis relies Financial analysis of Projected sector Financial projection Financial analysis of assessment on quantitative MSME MFIs (N=4) Banks (N=5) of MSME trends/indicators MSME owner assessment MFIs (N=4) Banks (N=5) For the processing of loan application procedures and approval, basic documents such as certificates of registration and single business permits are mandatory for most of For For the processing of loan application procedures and approval, basic documents the the processing financial application of loanBanks institutions. procedures also demand Tax and approval, Identification basic documents Number/ such Value Added assuch as certificates certificates of registration of registration and and single single business business permits are mandatory for most of theof Tax (TIN/VAT) registration certificates. Otherpermits are mandatory documents for most such as audited financial Banks Banks the financial institutions. financial also demand Tax Identification Number/ Value Added reports, institutions. matching collateral also demand and Tax Identification ownership certificates, Number/ Value certificates of Added asset, Tax (TIN/ guarantee Tax (TIN/VAT) registration certificates. Other documents such as audited VAT) registration certificates. Other documents such as audited financial reports, matching financial reports, matching collateral and ownershipcollateral and ownership certificates, certificates, certificates certificates of asset, guarantee of asset, letters guarantee and marriage 67 certificates might also be required for final processing of loan applications. 67 4.4.2. MSME Specific Products and Marketing Products. Regarding the product offering for the target SME market, the banks interviewed reported more than 80 percent of their loan products being term loans and other top loans include overdraft, pre-shipment credit and advances on import bills. MFIs provide group lending as their main loan product. They also provide non-financial products such 50 as trainings, technical assistance and services aiming at increasing market linkages to MSMEs. The main savings products offered to SMEs are saving and time deposit accounts for both banks and MFIs. Checking accounts are only offered by banks. Table 12 below summarizes the main products and services offered by MFIs and banks. Table 12. Product and Services Offered, by MFIs and Banks MFIs Banks Typology Scaled up versions of retail or microfinance Scaled down version of products offered to products corporates Product offering Loans for investment and trade, mainly via Working capital, loan for investment, trade a group lending model. Loans for leasing of financing and project financing are offered. agricultural equipment. No leasing and factoring credit products. Other non-lending products such as payroll/ pension retirement payment, other payment services and foreign exchange. Free business and technical trainings, Free payment services and free business and coaching to SMEs technical trainings, coaching to SMEs Saving products are time deposit accounts Saving products are cheque accounts, time and savings accounts. Compulsory saving deposit accounts and saving accounts. model is used in MFIs Network outreach Only regional. Focus on specific regions as National. No specific geographic focus. One affiliated with regional governments’ bank focuses on export Financial institutions in this survey do not seem to offer a sufficiently large product mix. Three banks reported offering either one or two of the following products: payroll/pension retirement payments, other payment services and/or foreign exchange. One bank offers all these three products while only one MFI reported offering payroll/pension retirement payments and other payment services. Other lending products such as leasing and factoring are not offered by any of the financial institutions in this study. Furthermore, products provided to the SME and micro enterprise market are largely standardized and efforts to continuously adapt them to client’s needs are limited. 88 percent of financial institutions in the study reported no change in their financial product offering between the years 2010 and 2012. This standardization, limited availability and range of products indicate that there is only limited lending product innovation taking place in the financial sector. Marketing. Table 13 below summarizes the main marketing strategies reported by interviewed banks and MFIs Table 13. Target markets of banks and MFIs MFIs Banks Depth of target market Micro enterprises and SMEs Large enterprises and some SMEs Target markets Mainly targeting micro enterprises with Mainly targeting large enterprises with very limited outreach to SMEs limited outreach to SMEs Agriculture and manufacturing are the Also agriculture and manufacturing (agro main focus sectors (as encouraged by the industry, export oriented) as the main focus government). sectors. SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 51 Banks and MFIs adapt their marketing efforts based mainly on profitability expectations, credit quality and specific industry sectors. There is a slight difference in sector targeting between banks and MFIs. In line with the government’s priorities, both banks and MFIs heavily target the manufacturing sector (80 percent of MFIs and 83 percent of banks). Banks’ overall involvement is highest in the manufacturing and export trade sectors, while lower in agriculture and construction. 80 percent of MFIs are marketing to target the agricultural sector compared to 67 percent of banks. As to the construction sector, 50 percent of banks focus on it and 40 percent of MFIs. The only sector banks are not targeting but 40 percent of MFIs are, is petty trade. MFIs on the other hand are not involved in export trade financing at all while 83% of banks are. Figure 31. Marketing of MFIs and Banks Focus MFIs Banks Figure 31. Marketing Figure 31. Focus of Marketing Focus of MFIs and and Banks 6 Marketing sector-specific Marketing sector-specific focus focus 6 Marketing Marketing Sectors Sectors SMEs for SMEs for 5 5 6   6   4 4 5   5   4   3 3 4   3   3   2 2 2   2   1 1 1   1   0   0 0 0   Manufacturing (Export) Agriculture Trade (Export) Agriculture Construction Yes   Yes   No   No   Manufacturing Trade Construction MFIs   MFIs  ((N=4)   N=4)   Banks  ( Banks   (N=6)   N=6)   MFIs (N=4) MFIs (N=4) Banks Banks (N=6) (N=6) Besides Besides sector sector specific Besides sector thetargeting, specific targeting, specific questionnaire targeting, the questionnaire the also explored geographic questionnaire also also explored explored geographic targeting and outreach. geographic targeting Geographic and outreach. location does Geographic not appear location to be an does importantnot appear to marketing be an important criterion targeting and outreach. Geographic location does not appear to be an important marketing for marketing financial institutions. criterion criterion for financial institutions. This is particularly the case for MFIs as the five of total This is for financial particularly the caseinstitutions. for MFIs as This the is five particularly dominant the ones, case with for more MFIs than as 90 the five percent dominant asset portfolio dominant ones, with of the ones, more micro with more than 90 than financing percent 90sector, percent areof total total asset ofaffiliated asset portfolio with the regional portfolio of of the micro financing micro thegovernments in Ethiopia. financing defaultare By sector, sector, are they providewith affiliated affiliated with the the regional financial regional services governments governments in these specificin Ethiopia. in Ethiopia. By default By geographical default areas they they provide provide in which they were financial services in services financial Regarding established. these intheir these specific specific geographical geographic geographical outreach and areas areas in in be as can seenthey which which they were were in Figure established. established. 32, most of the banks andRegarding Regarding their geographic their their geographic MFIs use only outreach outreachas own branches and and as can be as can distribution seen seen in in Figure be channels. Figure Some 32, 32, most mostand banks of of the the MFIsbanks banks use Point of and and MFIs use MFIs(PoS) only use only their their own branches own branches as distribution as distribution channels. channels. Some banks Some banks and and MFIs MFIs (ATMs) Sale devices for their transactions whereas only banks have Automated Teller Machines use use Point Point of Sale of Sale devices (PoS) devices (PoS) for their for and their transactions transactions whereas whereas only banks only banks have have with a very limited geographical coverage only MFIs use agents as service points. Automated Teller Machines (ATMs) with Automated Teller Machines (ATMs) with a very limited geographical coverage and only a very limited geographical coverage and only MFIs use agents MFIs use as service agents as service points. points. Figure 32. Banks & MFIs distribution channels Figure 32. Figure Banks & 32. Banks MFIs distribution & MFIs distribution channels channels 6 6 5 5 5 5 4 4 4 4 3 3 3 3 2 2 2 2 1 1 1 1 1 1 0 0 0 0 0 0 Own Branches Own Branches ATMs ATMs PoS PoS Agents Agents MFIs MFIs (N=4) (N=4) Banks Banks (N=5) (N=5) 70 70 52 5. INTERNATIONAL BEST PRACTICES IN SUPPORTING SME FINANCE The previous two chapters provided an in-depth analysis of constraints to access to SME finance as one of the key drivers for promoting job creation. Building on those findings, this chapter has three main objectives: To look at the main findings emerging from the demand-side and supply-side analysis and put them in relation to international best practices in SME finance; To provide an overview of the initiatives that the Government of Ethiopia has currently put in place to support MSMEs, highlighting whether alternative sources of finance are sufficiently exploited; To present how other countries (namely Turkey and China) have been successful in tackling the missing middle issue by implementing effective interventions in support of SME finance. 5.1. Key Findings on SME Finance Practices in Ethiopia vis-à-vis International Best Practices The demand-side and supply-side analysis conducted in the previous chapter provided a comprehensive list of findings and constraints that, if properly addressed, could unlock the potential of MSME finance for Ethiopia. In this section we will focus on what we consider to be the five main findings emerged from the analysis done so far: 1. The lack of a common definition of MSMEs across financial institutions and private sector operators 2. The effective existence of a missing middle phenomenon that constraints growth opportunities for small and medium enterprises 3. The excessive collateral requirements 4. The role of Government programs 5. The inadequateness of business models. 5.1.1. The importance of a commonly agreed definition of MSMEs Having a common MSME definition at the national level would ease the design of loans, investments, grants and statistical research. Worldwide, efforts to support MSMEs are at the center of the development agenda. Since the G-20 summit in Pittsburgh in 2009 the MSMEs opened a debate on whether a universal definition of MSMEs could be found. Hypothetically, the choice of an MSME definition could depend on many factors, such as business culture, the size of the country’s population; industry; and the level of international economic integration. A research recently conducted by IFC demonstrated the wide range of approaches government can take to define what an MSME is in their economy (Table 14). SME FINANCE IN ETHIOPIA: ADDRESSING THE MISSING MIDDLE CHALLENGE 53 Table 14. Selected MSME definitions across the world MSME Definitions Country Name Year (number of employees, unless otherwise noted) Micro Small Medium China 2009 new classification, which <300 Indstr. <2000 Indstr. might include micro will <600 Constr. <3000 Constr. take effect at the end of <100 Wholesale <200 Wholesale 2010 or 2011 <100 Retail <500 Retail <500 Transp. <3000 Transp. <400 Post <1000 Post <400 in Hotels & Restaurants <800 in Hotels & Restaurants Kenya 2006 1-10 11-50 51-100 EU <10; turnover<=≤ € 1 m <50; turnover<=≤ €10 m <250; turnover<=≤ € 50 m South Africa 2004 Agri <10, other <20; <50; turnover