EXECUTIVE SUMMARY Background, methodology and conceptual issues

1. As the microfinance industry matures, ownership and governance issues in microfinance institutions have
obtained an increasingly important role in the discussions about the conditions for success, or reasons for failure, of such institutions. In this context, the purpose of this Ownership and Governance Study is to study, establish and recommend options for efficient and effective governance of MFIs in Ethiopia. These issues have gained in importance as a result of the global financial crisis which has resulted in a worsening economic environment in Ethiopia for the operations of MFIs on the one hand, the demand for microcredit is likely to increase, but on the other hand it is getting increasingly difficult for MFIs to refinance their operations.

2. The Rural Financial Intermediation Programme (RUFIP) Coordination and Management Unit in the
Development Bank of Ethiopia (DBE), with funding being provided by the African Development Bank (AfDB), has awarded a contract to prepare the Study to BKP Development Research & Consulting GmbH of Munich, Germany. The study has been prepared over the period December 2008 to July 2009.

3. The study is structured as follows: Chapter 2 defines the key concepts and introduces the analytical
framework of the study. Chapter 3 presents international MFI ownership and governance (best) practices, and derives some first issues and recommendations for the microfinance sector in Ethiopia. Chapter 4 provides a brief overview of the MFI sector in Ethiopia. It is followed, in chapter 5, by an analysis of the Ethiopian legal and regulatory framework for ownership and governance of MFIs. Chapter 6 presents the details of our analysis of ownership and governance practices in Ethiopian MFIs. The final two chapters summarise the findings and recommendations and propose an action plan for the further strengthening of ownership and corporate governance practices of MFIs in Ethiopia.

4. The study builds on the already existing body of work regarding ownership and governance of MFIs in
Ethiopia as well as international experience, while adding new data and analysis obtained from original research. In order to analyse ownership and governance practices in Ethiopia, more than 50 interviews of MFI management, board members and shareholders of 13 MFIs (covering more than 60% each of total assets and capital in the Ethiopian microfinance industry) have been held.

5. The analysis of ownership and governance of MFIs in Ethiopia is based on the identification of the key
stakeholders for ownership and governance, their functioning and their relations with each other. First, the structure of ownership has an impact on governance as well as a direct impact on the performance of the MFI. With regard to governance in a strict sense, within an MFI the owners, the board and the management are the key stakeholders, with the board being at the centre. The analysis of governance thus needs to look at the functioning of the board its responsibilities, structure and procedures as well as its relations with, on

the hand, the owners, and management on the other. Finally, governance factors determined by the regulatory environment influence the MFI. There are thus five analytical dimensions to be distinguished. Key findings and recommendations regarding ownership

6. According to the Ethiopian law, MFIs must be organised as shareholder companies, i.e. 100% of
Ethiopian MFIs are shareholder companies. Internationally, this share is lower although shareholder firms constitute the most important type of MFIs. However, in view of the international empirical findings and also the self-perception of NGO MFIs the priority given to shareholder MFIs is debatable. Based on performance data, no ownership type is clearly superior to the others. This means that NGO owned MFIs constitute a useful complement to regulated MFIs especially in an economic environment where there is high excess demand for microfinance, i.e. where outreach is low and demand for microcredit cannot be satisfied. This clearly is the case in Ethiopia. Therefore, based on the international experience it should be considered to allow NGOs to run microcredit operations

7. Based on the international experience, a limited number of MFI owners (but more than two) is the typical
and preferable choice. In Ethiopia, the core capital sources of most Ethiopian MFIs are (foreign) NGOs or local governments, with ownership by individual persons also occurring in some MFIs. In the NGO and local government affiliated MFIs, economic ownership usually rests with one or two owners. However, the legal requirement for the establishment of MFIs as share companies has led to creation of nominal shareholders, because share companies must have at least five shareholders according to the Commercial Code. Hence, only to fulfil the legal requirement, mother NGOs or governments divide their original capital among at least four or five different persons (employees in some cases) or associations and establish the share company. These same people sit as directors but have no incentive to perform properly. This "grey" practice is commonly known and accepted but still the legal provisions should be amended in line with the practice. Otherwise MFI governance violates one of the most basic principles of governance, i.e. transparency, because it is never clear for outsiders if a shareholder is a "real" or "proxy"/nominal shareholder.

8. Foreign investors in MFIs are a common practice internationally. This is not the case in Ethiopia, where
the provision of financial services is closed for foreigners, for various reasons. The arguments most often used are, in addition to foreign exchange risks, that local financial services providers must be protected against stiff foreign competition (the infant industry argument), that foreign financial services providers are likely to concentrate on profitable business, thereby threatening to destroy universal service provision of financial services. None of these arguments appear to be valid in the microfinance sector, however: Given that the main target of MFI operations are the poor, there is very limited opportunity for MFIs to deal in foreign currency. Likewise, re-patriation of profits is not an issue for MFIs. Thus, the foreign exchange risk of allowing foreign investors is low.

MFIs, by definition, do not focus on the most profitable business but pursue a social objective, i.e. poverty alleviation. There is thus no cherry picking risk. The risk of capital flight could be further reduced by making the status of such foreigners as domestic investors who cannot repatriate their dividend in foreign currency out of the country.

9. At the same, time various arguments can be given in favour of allowing foreign capital participation in
MFIs: Formalising the "real" ownership structure in MFIs will have a positive impact on MFI governance structures when licensed foreign stakeholders actively participate in the governance activities, rather than through board members representing nominal shareholders. If MFIs established in the above way are to threaten and out-compete MFIs formed by local sources, the regulator could take due legal measures in due course to restrict either the capital or outreach of the former MFI. The positive externality of having competitive foreign MFI operators with domestic investor status should also be underscored. With the coming into force of the new Civil Societies and Organizations law, many foreign NGOs will not be allowed to work in democratic rights, governance, and humanitarian issues. Allowing foreigners to participate in economic activities, such as MFIs then could give a guiding role in reallocating resources and raise the entrepreneurship skill of the poor.

10. The identification of options for Ethiopian MFIs to raise additional funds will become increasingly
important in view of rising inflation and the domestic impact of the global financial crisis. In the course of the study various options have been identified.

11. Deposit and savings mobilisation by MFIs: All MFIs established in accordance with the provisions of the MFI
Proclamation (No. 40/1996 now repealed and replaced by the Micro Financing Business Proclamation, Procl. No 626/2009) are legally authorised to accept deposits. However, even after accounting for age, the MFIs affiliated to regional governments seem to be doing better than other MFIs. The five largest MFIs account for 95.4% of deposits. One area of policy intervention to consider in this respect is the introduction of deposit insurance covering all deposit taking institutions. While this raises the cost to MFIs, the knowledge that their deposits with all MFIs have insurance cover may attract depositors. Besides, the high rate of inflation, if not brought under control, may be expected to pose further challenge to deposit mobilisation by MFIs because: (a) it means the real interest rate on deposits is highly negative discouraging savings; and (b) it erodes the saving capacity of potential depositors.

12. Government-NGO partnership in MFIs. Another alternative to consider is for foreign NGOs to get into
partnership with regional governments in establishing and expanding MFIs, since they both are social investors (in that both have social returns as key objectives), hence have more compatible objectives. They are also partners in the poverty alleviation objective. To our knowledge to date partnerships exist only with

local NGOs. Foreign NGOs are not involved as shareholders in the government affiliated MFIs. Nor are regional governments participating in the NGO promoted MFIs. It is not clear why this has not been done up to now.

13. Raising funds for MFIs through Government/Municipal bonds. Another option is borrowing which, in the
Ethiopian context, would mainly mean domestic bank loans (foreign loans, even from multilateral organisations, are unlikely to be forthcoming without government guarantee of both the principal as well as foreign exchange availability). However, experience so far shows that the commercial banks, private banks in particular, are not particularly keen to lend to MFIs (although there are some instances), more so with longterm loans. So, there is a need to consider alternative ways of raising funds, long-term funds in particular. While it would be difficult to imagine the issuance of MFI bonds, Government could also consider raising long-term capital through the issue of long-term government bonds and pass the funds to MFIs for on lending. This may be a better alternative to bank loans for at least two reasons: (a) it allows the MFIs to expand without being burdened with short-term repayment obligations of the principal; and (b) the borrowing interest rate is bound to be lower as the borrower is the government, presumed to be a safe borrower.

14. Allowing foreign social investors/donor participation. It appears that the ownership related problem needs to be
resolved if the financial capacity problems of MFIs are to improve. One way of doing this would be to allow social investors, including foreign ones, into the micro finance (and micro insurance) sector. Policy makers may be concerned about the capital account and foreign exchange availability implications of permitting this. However, they could start with permitting inflow of such funds that do not involve repatriation of profits and proceeds from the realisation of assets.

15. IFC-Private sector partnership in MFIs. The key purposes of the International Finance Corporation (IFC) is
to create opportunity for people to escape poverty and improve their lives by, among other things, helping to generate productive jobs and deliver essential services to the underserved. This is quite consistent with what MFIs do. The IFC gets involved at the firm-level through direct investments as equity partner and through advisory services. It normally provides funding for new projects (up to 35% of project cost) as well as expansion (up to 50% of project cost) subject to limit of 25% of the total capitalisation of the company) through equity participation with the private sector. It also shares the risk fully. It may thus be useful for some of the existing private MFIs (and new ones) to explore the possibility of getting such equity funding from the IFC. Such arrangement allows the MFIs (a) to access funding from a source which also fully shares the risks; and (b) to benefit from the advising services of the IFC as well as its monitoring and follow up as equity investor.

16. Accessing the Microfinance Enhancement Facility of the IFC. Cognizant of the effect of the global crisis on
microfinance institutions around the world, the IFC and the German development bank, KfW, have recently launched a new facility called Microfinance Enhancement Facility to assist MFIs. It is a short- to medium-

term refinancing facility with, expected to reach $500 million and intends to provide refinancing to more than 100 MFIs. Relevant government institutions (e.g. the National Bank of Ethiopia) and AEMFI should take appropriate steps to enable Ethiopian MFIs take advantage of the Facility.

17. The different options outlined above need to be pursued simultaneously. The effect of inflation is bound
to be common for all MFIs, small or large, although it may be more intense among the young and small MFIs. After all, given that existing MFIs differ in ownership/ affiliation, the above options may not apply to all MFIs. For example, IFC partnership may be expected to be more relevant for MFIs with real private owners. The option of raising funds through government bonds and passing it on (together with the debt) to MFIs may be appropriate for those that have the capacity to meet the repayment obligation. On the other hand, foreign social investor participation may be more appropriate for existing young and small & medium MFIs as well as for establishing new ones. The Microfinance Enhancement Facility of the IFC is applicable to MFIs facing debt refinancing problems. Key findings and recommendations regarding MFI governance

18. A number of gaps in the governance practices of Ethiopian MFIs have been identified in several of the
five analytical dimensions, and recommendations to overcome these have been developed.

19. Board size. The number of board members in surveyed Ethiopian MFIs ranged between 3 and 9, with
majority of the MFIs having 7 board members and the average being 6.25. Boards are thus smaller than in other countries. Although empirically, no clear link can be established between board size and MFI performance, boards which are too small risk overburdening individual board members and do not facilitate the creation of board committees (see below). Furthermore, larger boards simplify easy transition from year to year. It is therefore recommended that MFIs with board consisting of fewer than seven members increase board size to a minimum of seven board members. The number of board members should be uneven in order to prevent impasse situations. An increase in board size would ideally be done by adding independent board members, although this will require a change in law.

20. Board composition and election. Normally, the board is elected by the general assembly. However, the role of
the general assembly in many Ethiopian MFIs is strongly limited by two factors, i.e. the small number of shareholders and the over-involvement implicit in cases of nominal shareholders of regional government and NGOs. Regional governments and NGOs that have stakes in MFIs do not seem to have a clear policy on how they exercise their power in nominating board members. Thus, direct assignment of board members by some authorities has been observed. This is not good practice as it undermines the basic rules of the institution. The board needs to be elected by those people who are assigned to represent the institution as the shareholder. A means of addressing the above issues and at the same time improve corporate governance at least in line

with international empirical findings and best practice recommendations is to include independent members in MFI boards, i.e. individuals who are neither owners nor other stakeholders (employees, clients, etc.). Such practice is commonly used in other countries and has been shown to improve MFI performance both in relation to social and financial objectives.

21. Board qualifications. When compared to international experience where experience in financial
analysis/accounting is the foremost requirement the qualification of Ethiopian MFI boards could still be enhanced. Likewise, the number of board members that have the required understanding of the microfinance business is considered insufficient. This is directly related with the inadequacy or complete absence of training opportunities for board of directors provided in most MFIs. This significantly limits the effectiveness of the board and may lead to the over dependence of board members on the management. In response, the development and implementation of compulsory training in financial analysis and microfinance for board members without a related degree should be considered to broaden the knowledge base of board members and enable them to more effectively monitor the performance of their institutions. Such training, complemented by other courses deemed relevant, could be provided by AEMFI (which is already facilitating board training), the Micro and Small Enterprise Development Agency or other suitable institutions.

22. Board tenure. The tenure of board members which is either 2 or 3 years seems appropriate. However, the
concern regarding tenure is related with the number of re-elections that board members can have. In some cases re-election is a must due to the low number of shareholders. Also, founding documents often neither restrict the number of times that one can be re-elected nor establish overall term limits. International practice has clearly shown that rotation of board members should be part of the game and maximum term limits should be established in order to prevent collusion between board and management. This can be realised through a variety of measures, one of which is to make amendments to the founding documents and establish term limits.

23. Board committees. International experience has shown that board committees facilitate effective
understanding of issues at board level. They also enhance the quality and quantity of board decisions. Hence MFI boards should have a few permanent committees that deal with issues such as risk management and strategic planning and a few other ad hoc committees on emerging issues from time to time. The nonexistence of board committees among Ethiopian MFIs a serious bottleneck for the effectiveness of boards and hence needs speedy resolution. The establishment of committees may also require some MFIs to extend the size of their boards seven board members can be considered to be the minimum size for the establishment of committees.

24. Board documents. Ethiopian MFI boards mostly rely on the institution's founding documents for their
board activities and these are insufficient and inappropriate. Specific guidelines that stipulate board duties and responsibilities and a range of other relevant issues should be available with all MFIs. The effectiveness of the board without such documents is significantly hampered. Besides, the existence of a code of ethics to which

all board members must subscribe can set a high moral ground for all other internal and external stakeholders of MFIs.

25. Conflict of interest. As financial institutions MFIs need to protect their board of directors and senior
management members from committing mistakes emanating from conflicts of interest. Where there is no such protection deliberate and non-deliberate instances of conflict of interest may occur and institutions will remain unprotected. It is hard to state whether there have been instances of conflict of interest in Ethiopia or not. First of all, the issue of what conflict of interest is not clear. Most Ethiopian MFIs do not have a detailed policy on conflicts of interest. This is in stark contract to international practice and recommendations. The presence of strong rules and procedures for the avoidance and handling of conflicts of interest is one of the bases of good corporate governance. The definition and introduction of such codes in Ethiopian MFIs is therefore considered essential.

26. Board performance and Evaluation. Another component of good corporate governance is to measure
performance of the supervising body, i.e. the board. This can be done through self-evaluation (which is the more common practice in MFIs) or hiring external evaluators. In Ethiopia, no MFI board has the habit of formally evaluating itself. In fact, even the evaluation of CEO/management performance is often not performed by boards. Since measuring the people at the top sets a very high standard of commitment for the success of the institution this gap needs to be addressed as soon as possible. MFI Ownership and Governance Action Plan

27. The recommendations have been further developed into an "Ethiopian MFI Ownership and Governance
Action Plan" which is suggested to be jointly implemented by all stakeholders of microfinance in Ethiopia. The plan distinguishes actions related to MFI ownership and sourcing additional funding, actions related to the improvement of MFI governance, and miscellaneous other measures.