WPS5141 Policy Research Working Paper 5141 State Financial Institutions Mandates, Governance, and Beyond Heinz P. Rudolph The World Bank Financial and Private Sector Development Vice Presidency Financial Systems Department November 2009 Policy Research Working Paper 5141 Abstract There is no doubt that on average the performance of system would be preferable. This paper analyzes the state financial institutions around the world has been experience of four state financial institutions that have below the lowest expectations. Lack of governance, performed relatively well in the past: Canada's Business management skills, regulation, and transparency, and Development Bank, Chile's BancoEstado, South Africa's misguided incentives have contributed to discredit Development Bank of Southern Africa, and Finland's these institutions for supporting the development of Finnvera plc. The author finds that these institutions local financial markets. However, the pro-active role have different checks and balances to mitigate eventual that some state financial institutions have played in the mismanagement of resources. The author also finds that recent crisis in allocating credit to sectors cyclically not little progress has been made in measuring the policy attractive for commercial banks has brought back the performance of these institutions. question of whether some state ownership in the banking This paper--a product of the Financial Systems Department, Financial and Private Sector Development Vice Presidency--is part of a larger effort in the department to strengthen financial systems by improving the quality and effectiveness of state interventions in finance.. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The author may be contacted at hrudolph@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team State Financial Institutions: Mandates, Governance, and Beyond By Heinz P. Rudolph The World Bank State Financial Institutions: Mandates, Governance, and Beyond By Heinz P. Rudolph1 The World Bank 1. Introduction In the past three decades the presence of State Financial Institutions (SFIs) has been highly controversial. The evidence suggests that state-owned banks have underperformed and have negatively impacted economic growth. La Porta et al. (2002) find that government ownership of banks is associated with slower subsequent financial development, lower economic growth and lower growth of productivity. Beck and Levine (2002) fail to find any positive effect of government ownership of banks on growth. Caprio and Martinez Peria (2002) show that government ownership of banks is associated with a higher likelihood of banking crises. Dinc (2005) provides evidence that state financial institutions increase their lending during an election year and that in emerging markets state financial institutions finance the government to a greater degree than do private banks. Most of the poor performance of SFIs is related to the lack of a clear mandate, and a governance system that allows the presence of weak boards of directors and management, which are subject to political intervention. Evidence provided by Caprio et al. (2004), La Porta et al. (2002), and Dinc (2005) shows that most state financial institutions around the world are characterized by lack of managerial skills and subject to government intervention in credit decisions. This paper discusses good practices for aligning the functions of SFIs with solving market failures. Mismanagement of resources is frequently a consequence of weak board of directors and management, political interference, and lack of a clear mission of the SFI. Based on four case studies, this paper analyzes the importance of having clear and sustainable mandates, and an institutional framework able to mitigate political interference. The paper analyzes good practices 11 The author is a Senior Financial Sector Specialist in the Financial and Private Sector Development Vice- Presidency of the World Bank (hrudolph@worldbank.org). The author would like to thank senior management at BDC, BancoEstado, Finnvera, and DBSA for providing insightful information for this note; David Scott, Jerome Nycz, Lily Chu, and Francisco Bernasconi for useful comments; and Denisa Mendelsohn for valuable research assistance. The paper is part of a larger effort in the department to strengthen financial systems by improving the quality and effectiveness of state interventions in finance. The views presented in this paper do not necessarily represent the views of the World Bank. 2 for building, implementing, financing, and evaluating the mandates of SFIs, and examines the experiences of these SFIs regarding the ownership and nomination of board members. The analysis is based on four case studies of wholly state owned banks: Canada's Business Development Bank (BDC), Chile's BancoEstado, South Africa's Development Bank of Southern Africa (DBSA), and Finland's Finnvera plc. While attempting to diversify the geographical pool of our case studies, we selected these four state owned banks because they, in our opinion, operate successfully within their predetermined set of targets, exist in a sound functioning regulatory framework, and can be regarded as benchmarks for other state financial institutions. Three of the four banks are development banks, and BancoEstado is the only commercial bank, but with a social objective. BDC, a Canadian development bank, is the only state owned bank in Canada. It provides support to Canadian entrepreneurship, with special emphasis on SMEs. While lending to Canadian businesses is the main area of business, BDC is also involved in venture capital, consulting, and subordinate financing. BancoEstado is the third largest commercial bank in Chile and also the only government owned bank. Although it is increasingly evolving towards a universal bank, its main social focuses are supporting access to financial services for poor or geographically isolated individuals, lending to micro- and small businesses, and housing financing for low income people. DBSA is one of the four state owned development banks in South Africa. Its main objective is to facilitate financing and technical assistance to companies and municipalities with a focus on infrastructure funding. It also provides financing to large infrastructure projects in neighboring countries. DBSA taxable profits go to DBSA Development Fund, a subsidiary of DBSA, main objective of which is to provide support to low income municipalities through grants, deployment of technical expertise, training and identification of projects. Finnvera plc, the only state owned financial institution in Finland, provides funding to SMEs, and promotes internationalization of the Finnish companies. Finnvera operates also as an export credit agency (ECA). The paper highlights the need for defining clear and sustainable mandates aimed at solving a market failure. Both BDC and Finnvera have clear and sustainable mandates and play a role that is complementary to that of privately owned commercial banks. As the mandates of BancoEstado and DBSA are less clearly defined, they get more engaged in competing with privately owned commercial banks. The Corporate Plan of SFIs is a useful tool that owners can use to focus and monitor their mandates. In the case of DBSA, for example, the Corporate Plan imposes targets that emphasize the development impact of the bank over the commercial aspects of the operations. The paper emphasizes the importance of self-financing of the mandates and limiting the distortions on the rest of the financial sector. All of the SFIS analyzed fund themselves from the market, which has helped to impose market discipline on the management. High quality of management and boards of directors are one of the common elements of the four cases analyzed. While competent board of directors and management in the cases of BDC and 3 DBSA are a consequence of a transparent mechanism to bring independent directors to the bank, in the cases of Finnvera and to a less extent BancoEstado, they are a result of cultural factors. In all four cases, compensation to management, although lower than in privately owned commercial banks, is better than in the public sector.2 A structure of clear ownership of these institutions is also analyzed. The four SFIs have an identifiable owner inside the government, which in most of the cases monitors closely the activities of the SFI. The paper is separated into four sections, including this introduction. The second section discusses the mandate, in particular how to define a mandate, and the different experiences in implementing a mandate and funding the operations. The section also analyzes the indicators for measuring the public policy performance of SFIs. The third section discusses governance issues, putting particular attention on the ownership policies, the process of nomination of the board members and the professional qualifications of the management. The last section concludes. 2. The Mandate of SFIs 2.1 Defining the Mandate The inability to define their objectives is one of the main problems of SFIs. Most countries have great difficulty in justifying the presence of SFIs. In many cases, these SFIs are a legacy from past administrations that retain the political power to stop any attempts of reforming them. In other cases, it is in the interest of the government to maintain SFIs since it can leverage subsidies to groups of interest through the SFI. The SFI can simultaneously finance operations that have nothing to do with the objectives of the bank with cross subsidies from profitable sectors of the SFI. In order to be able to gain market share with these profitable businesses, SFIs offer rates below the market, which ends up not only crowding out the role of privately owned institutions, but creating large and fragile SFIs that are highly dependent on government support. Before creating an SFI, the government needs to identify the market failure that it is trying to solve. Market failures are usually associated with a lack of financial support to certain groups of the population, for example in the agricultural sector, infrastructure sector or SMEs. A market failure can also exist on the liability side of the bank, more precisely caused by a lack of saving instruments for certain groups of the population. Once the market failure that requires resolving has been defined, the government needs to define the mechanisms to deal with the problem. While in the past banks were one of the few alternatives, currently there are a broader set of alternatives that do not necessarily require the creation of a SFI. For example, weather related risks are a common impediment for the access to credit by farmers. While in the past the creation of agricultural state owned bank was seen as the 2 Public sector wages in Finland are not substantially different than market levels. 4 solution to this problem, the use of more innovative solutions, including the use of catastrophe insurance, might be able to diversify the weather related risk towards other investors and facilitate the interest of commercial banks in lending to farmers. Another example is in the area of infrastructure. The lack of long-term funding sources is a common impediment for infrastructure financing. Banks may see high risks in investing in long term projects. While in the past the creation of state owned bank was seen as the solution to this problem, an active role of governments in developing an interest rate yield curve with government instruments may facilitate the interest of institutional investors in financing these long term investment projects. These pro-market developments provide targeted solutions to market failures and promote competition in the credit market. However, it might be the case that an SFI is the most efficient solution. Even in developed economies, funding for venture capital and SMEs is still challenging. In most of the cases, the market fails to value the risk of these projects and provide limited funding. This opens the room for a more active involvement of the government in promoting financing to venture capital and SMEs. Some governments provide this support in a decentralized way (e.g. SBA in the USA), while others do it through development banks (e.g. BDC in Canada). It is not our intention to show that one model is better than the others, but to show that SFIs can be a viable solution under certain circumstances.3 In order to focus the activities of SFIs, their mandates should be stated in the law. Since SFIs are created by law, the mandate should also be stated in the law. Stating the mandate in the law has important benefits, especially for focusing the activities of the SFI. As the board and management of the SFIs become accountable for fulfilling the mandate, they become less open to pressures and involved in business that is the province of the private sector. Based on the experience of the four cases, we propose that SFIs mandates should include at least three components: a. A target sector b. Rules of cooperation with the private sector c. Minimum level of efficiency a. Target sector The target sector needs be justified by a market failure. Typical market failures can be found in business that are considered to be too risky for commercial banks, including micro and SME lending, the agricultural sector, and sectors that require long term funding, such as infrastructure 3 This argument is consistent with the pro-market activism views proposed by De la Torre el al (2007) 5 and mortgages. Since market failures are dynamic over time, governments need to review the mandates to periodically ensure that the market is unable to fulfill these functions.4 In order to focus the attention of directors and managers, the mandate of SFIs should be clear and specific, but should allow for flexibility for the SFIs to tailor its financial products to the market needs. Although broad mandates allow SFIs to diversify their portfolios better, they tend to create confusion about the real objectives of the SFI, opening room for cross subsidization, inefficiencies, and hidden operating losses. A board of directors should have the flexibility to operate with different products to fulfill its mandate, including the use of loans, guarantees, securitized instruments and derivatives.5 However, the coverage of the target sector needs to be sizable enough to create a sustainable business for the SFI. Some countries create SFIs with the purpose of servicing specific sectors of the economy are not sufficient to ensure a sustainable business for the SFI over time. The most common examples are SFIs involved in lending to farmers, in areas with scarce technical innovation, which ultimately become unprofitable for the SFI. If the target sector is not profitable, the SFIs became an institution that only leverages the subsidies from the government.6 Under these circumstances, the incentives for efficiency are blurred and the directors and managers lose their independence from the government. As discussed below, building a profitable business may require SFIs to build on targeted subsidies from the government, but they should avoid running an unprofitable business that may end up with losses and frequent recapitalizations. Mandates need to be revised on a regular basis and SFIs should disappear when the market need is no longer present. Self-perpetuation is one of the main problems of the SFIs. Governments should review and potentially revise the mandates at least once every decade in order to ensure that the market failure still exists and that the presence of the bank is still justified. This sound practice is followed by Canada's BDC. During a decade, it is likely that financial innovation may have created instruments to give access to sectors that, at the time when the Law was enacted, did not have access to suitable credit. 4 It is important to clarify that since the mandate is unclear, we do not see a permanent role of the state in the ownership of commercial banks. In most of the cases, these institutions can be replaced by more efficient private institutions. 5 This assumes that the SFI has the capacity to manage these risks and that the SFI is subject to prudential supervision. 6 The situation of unsustainable banks gets worse when they are allowed to take deposits from public, since the likely failure of the bank implies that the government has to take the losses, rescue the depositors and add capital to the bank. 6 b. Rules of cooperation with the private sector In order to avoid public intervention in areas that are properly the domain of the private sector, the mandate needs to establish that SFIs should play a complementary role to that provided by privately owned commercial banks. SFIs should complete markets not to act as a substitute for the role of private financiers. In this context, the mandates of SFIs should implicitly promote the participation of privately owned commercial banks. SFIs may support the market in various ways, for example by segmenting the market and servicing mainly clients that are not serviced by others, by cooperating with commercial banks to make viable projects that otherwise would not be financed, or simply by developing products that may complement the role of commercial banks (i.e. guarantees or derivative products). Although SFIs should avoid getting involved in business that is more properly the province of the private sector, under certain circumstances SFIs may want to keep some presence in these markets. De la Torre and Ize (2009) suggest that commercial banks suffer from cyclical changes in risk aversion, becoming more risk averse in periods of recession. This market failure may justify some limited but permanent presence of SFIs in markets that are serviced by the private sector, which might be affected by cyclical factors. If the role of the SFI is to serve as a buffer in the event of a contraction in credit, the SFI needs to understand the market and therefore some permanent presence is justified. SFIs will only be able to expand their credit portfolios quickly, if they have internal knowledge of the market. SFIs need to be existing market participants and insiders of the credit market in order to expand effectively their activities when needed. c. Efficiency The mandate should establish that the state expects some type of return on its capital. The mandate does not need to specify the rate of return, but it should say that the SFI needs to be financially sustainable over time. The requirement for financial sustainability is about creating the incentives for the board of directors and the management of the institution to make proper use of the available resources. If profits are not an objective, the administration of the SFI may have fewer incentives to conduct adequate risk management. Unprofitable SFIs that receive regular support from the government have incentives to hide operational inefficiencies in those losses. While in most markets efficiency can be attained through competition, in less competitive markets (such as those of SFIs) it is necessary to establish some minimum parameters of efficiency. With the exception of BancoEstado, the other analyzed banks look for a return that at least resembles the cost of funding of the state. As a commercial bank, BancoEstado requires a higher return on capital than the other banks. Some exposure to competition to SFIs that act as buffer to cyclical behavior of banks might be healthy, as it imposes an additional benchmark for 7 measuring performance. For example BancoEstado has benefited enormously from competition, but at the price of involving a state institution in business that should belong to the private sector. Comparative analysis In most of the cases, mandates in the four cases analyzed do not fulfill these three criteria. While the target sectors of BDC and Finnvera are relatively clear and specific, DBSA's target sector is subject to interpretation. Target sectors of BancoEstado are not defined in the law, but are based on the view of the management and supervisory board. While BDC and Finnvera do not compete with the private sector but cooperate with them in the structuring of financing for companies, DBSA and BancoEstado compete with private business to make money to serve sectors that are less serviced by commercial banks. Although financial sustainability criteria are not included in the mandate of the cases analyzed, all of them operate with minimum criteria of efficiency. Table 1 provides a description of the main elements of the mandates. Table 1. Structure of the Mandate Market Failure Competition Efficiency BDC SME BDC does not compete Minimum ROE should be with commercial banks, as equal to the Government it provides financing for long-term cost of capital higher risk transactions and prices for risk by security shortfall BancoEstado Broadly defined in the Compete with commercial Minimum ROE should be Law, but focuses on micro banks similar to other and SME lending, and commercial banks mortgage financing for low income people DBSA Municipal Infrastructure Competes with commercial Minimum ROE should be banks, but requires at least 10 percent minimum co-funding with other institutions Finnvera SME Cooperates with Minimum ROE should be commercial banks to at least equal to the six- facilitate financing to month Euribor rate companies Source: World Bank In the case of BDC, the mandate is defined as: "The purpose of the bank7 is to support Canadian entrepreneurship by providing financial and management services and by issuing securities or otherwise raising funds or capital in support of those services."..."In carrying out its activities, the Bank must give particular consideration to the needs of small 7 BDC 8 and medium sized businesses."..."The loans, investments, and guarantees are to fill out or complete services available from commercial financial institutions." (BDC Act of 1995) BDC's mandate has an important number of advantages: first, it is short and clear, defining the objective as supporting Canadian entrepreneurship, with particular consideration to the need of SMEs. Secondly, it defines the role of BDC as complementary to the commercial banks and mainly for the purpose of completing the market. Thirdly, it allows BDC to operate with multiple instruments (loans, investments and guarantees), which gives the Board and the management of the company the necessary flexibility to service its clients. Although it is not in its mandate, the government requires BDC to be commercially viable and to earn a return on equity at least equal to the Government's average long-term cost of capital. BDC has a high quality database with disaggregated data on Canadian SME companies since 1980, which together with state of the art risk management systems allow BDC to price by risk and to run a profitable business in segments of the market that are not serviced by commercial banks. BDC is generally more expensive than the market in funding companies / projects that have other sources of funding, but it is an attractive source of funding for riskier projects that are less attractive or unattractive to commercial banks. On the basis of rational decision making, borrowers with other sources of funding do not use BDC.8 Since pricing is the source of market selection, many companies initially funded with BDC look for refinancing with commercial banks at cheaper rates. Approximately 6.5 percent of BDC's clients fully prepay their loans every year (based on average of number of clients over the past ten years). While a commercial bank may consider this as a business failure as the model depends on keeping the good clients, BDC considers this a success since projects can get funding directly from the market and do not depend on official support. However, this imposes an important challenge on the BDC management to bring new clients and rebuild the portfolio of clients on a regular basis. As a term lender (with minimum term offering of five years), BDC's average portfolio term is approximately seven years. BDC has also implemented other mechanisms to avoid displacing the private sector. First, it works in partnership with all other Canadian financial institutions in support of Canadian SMEs in an effort to complement their services rather than compete with them. Secondly, BDC has invested in corporate research that regularly tracks changes in the SME market conditions and identifies needs of the Canadian SMEs that are not being serviced by other financial institutions. In this context, BDC examines the impact of BDC products and services on the SME market as well as on other financial institutions. BDC conducts SME research via its own internal resources, and also partners with external third parties (Industry Canada, CIRANO, and Statistics 8 BDC has a small group of some large clients as well, which voluntarily stay with DBC due to their longer term vision of the projects, in particular during stress episodes. BDC participates also with in pari passu lending. 9 Canada). The results of the research are incorporated into BDC's processes to better execute the mandate and avoid displacing private sector financial services activity. Finnvera's Act defines the mandate as follows: "... to provide financing services to promote and develop business, particularly that of small and medium-sized enterprises, and to promote and develop the exports and internationalization of enterprises. In its operations, the State-owned specialized financing company shall promote the State's regional policy goals. Operations shall focus on shortcomings in the supply of financing services". "The company pursues financing by granting and administering credits, guarantees and other contingent liabilities, and making equity investments. The company also carries out research and surveys related to business financing and provide enterprises with development assistance, services, and advice". (The Act on the State-Owned Specialized Financing Company 18 June 1998/443) The mandate of Finnvera is clearly articulated and touches on most of the relevant issues. First, it provides a clear definition of Finnvera's purpose: to promote and develop SME business, promote and develop the export sector, and to foster internationalization Finnish enterprises. While the scope of the sectors in which it is mandated to operate is relatively broad, the areas of business are clearly specified.9 Secondly, the mandate defines the role of Finnvera as complementary to the role of the financial sector ("shall focus on shortcomings" and "should supplement the financial market by providing financing to exports, SMEs and growth companies"). Thirdly, it gives Finnvera flexibility to operate with different instruments (credits, guarantees and other contingent liabilities) for servicing its mandate. Although the law does not specify the commercial purpose of Finnvera, the government requires a minimum return on its capital, which in practice is equivalent to six-month Euribor rate. Finnvera does not compete with commercial banks, but cooperates with them in order to make feasible certain deals that would otherwise fail without the presence of a lender with longer-term perspective and less risk aversion. Finnvera is rarely the only lender and typically participate in syndicated lending and provides funding and guarantees. In many cases, it takes junior participation in syndicated lending or first losses. Finnvera conducts rigorous risk assessments of the projects before participating in these deals, and the assessments are supported by risk management models and skilled employees and management. In the case of BancoEstado, the mandate is only broadly defined in the law: "the objective of the bank is to provide banking and financial services to support the development of the activities in 9 Finnvera is also an Export Credit Agency. 10 the country..." However, in our interviews with current and former members of the management board, we noticed a more specific common ground among them: "the mission of BancoEstado is to maximize a social function subject to having a rate of return similar to the rest of the banking industry." Although this statement is not written in any official document, it seems to reflect the implied mandate that the management board has followed in the last two decades.10 Although having an implicit mandate is an important shortcoming compared with having a mandate that is written in the law, this implicit mandate has certain advantages and shortcomings. Firstly, it does not specify the target sector, which opens room for the bank to develop multiple activities and to divert its attention from its primary purpose. Secondly, although the implicit mandate assumes that BancoEstado's target market is related to sectors that are not serviced by privately owned commercial bank, it fails to address the issue of competition with commercial banks. As discussed below, BancoEstado competes with other commercial banks in multiple areas of business. Thirdly, the implicit mandate imposes on the management a minimum requirement in terms of return on equity, which has been essential for bringing efficiency to the bank. BancoEstado competes with the private sector in multiple areas of business. The management of BancoEstado sees the competition as a way of improving efficiency and accomplishing the minimum returns that are required by the government. In our interviews, we observed that the current and past management of BancoEstado have seen competition as a mechanism for improving the efficiency of the bank and for creating expertise in the design of financial solutions to the less well serviced sectors of the economy. Since BancoEstado does not receive any subsidies from the government,11 it sees the need for accessing the profitable parts of the market, including lending to large companies, in order to generate resources for supporting its mandate and achieving its financial targets. The lack of subsidies from the state and the fact that BancoEstado plays on a level field makes it existence less controversial with privately owned commercial banks. In our interviews with current and former executives of BancoEstado, we noticed a firm defense of their participation in the businesses that are properly the purview of the private sector. They argued that it in Chile it would not be commercially feasible to run a niche bank that focuses exclusively on lending to low income households, microenterprises and SMEs. Some attempts, including those by Banco del Desarrollo in Chile, were unable to develop the scale and to bring the technology, management, and expertise that is necessary to run a business like this. The business model of BancoEstado requires all the investments to be profitable from the operational 10 The annual report articulates BancoEstado mission as: "to be a universal bank, of all and for all, offering integrated financial services so that every Chilean, everywhere, can undertake entrepreneurial activities and develop." In our interviews we did not notice that current and former senior management team had this mission internalized. 11 In the past, BancoEstado received an implicit subsidy since it managed all the banking accounts of the government. With double digit inflation levels, and a government unable to manage its assets actively, BancoEstado was receiving a substantial amount of resources. The reduction in the levels of inflation and the active cash management of government accounts reduced this source of funding to a level that is practically irrelevant. 11 perspective, but the activities in the underserviced sectors of the economy (e.g. micro and SME lending) would be unable to be profitable without infrastructure provided by the bank, including the network of branches, and risk management systems. BancoEstado representatives argue that these business areas became feasible to service once the bank had the technical and human infrastructure to operate in them. BancoEstado also see the competition as a powerful tool for efficiency, which would have not been present if the bank had focused only on the less competitive targeted sectors of the economy. DBSA's regulations define the bank's mandate as focusing on "the provision of financial, technical and other assistance to achieve the objectives of the bank, with a focus on its investment activities on infrastructure funding, broadly defined, and with the object of acting as a catalyst to maximize private sector access to opportunities in the provision of public funding." Firstly, DBSA's target sector is broadly defined, which leaves room for interpretation about the meaning of infrastructure funding. In fact, the management of the institution has a broad interpretation of the concept, the boundaries of which are not clear. In addition, DBSA has the objective of generating profits with the purpose of transferring the taxable revenues to support DBSA Development Fund. However, this arrangement might be conflicting with the objective of supporting infrastructure projects for middle income municipalities.12 The reason is that while DBSA Development Fund focuses almost exclusively on financing of the poorest municipalities and DBSA may be under constant pressure to generate enough profits and may, thus, be more motivated to provide financing to large municipalities that result in higher profits. We see high value added in the resources invested by DBSA Development Fund, as it allows these municipalities to build capacity to have access to the financial sector in the future, but the link between DBSA profits and DBSA Development Fund may justify more involvement of DBSA in private projects that are not related to the mandate. Since DBSA Development Fund is mainly a provider of subsidies, we do not see a reason for linking its funds to the performance of DBSA. Secondly, the DBSA mandate does not address the complementary nature of its activities, and therefore DBSA gets regularly involved in the financing of private infrastructure projects. With the exception of lending to medium size municipalities, which represents about 30 percent of the value of its investments in 2007/08, most of the lending of DBSA is in competitive markets, including lending to large municipalities, private infrastructure projects, and premium international operations.13 DBSA has minimum targets in terms of co-financing with private financiers in order to ensure a complementary role of DBSA. Thirdly, although it is not defined 12 DBSA Development Fund serves to unlock municipal infrastructure grants and other resources from the government through the deployment of technical expertise to under-skilled municipalities to accelerate the implementation of infrastructure projects, mainly in the areas of sanitation and water. 13 Premium international operations are typically large infrastructure projects in neighboring countries that have the guarantees from the respective governments. 12 in the mandate, the corporate plan adds some tension between conducting prudent banking and maximizing development impact. DBSA's target is a return on equity of 5 percent per annum.14 2.2 Implementing the Mandate It is not only necessary to have a clearly delimited mandate, but also to be able to implement it. Therefore, SFIs need to put in place mechanisms to bring mandates into the priorities of the company. Traditional mechanisms include the development of a corporate plan and some type of performance indicators for aligning the business strategy with the mandate of the company. Table 2 Main Highlights in Corporate Plans BDC BancoEstado DBSA Finnvera Client Satisfaction Financial Results Development Impact Minimum amount of guarantees and loans Employment Engagement Client Satisfaction Sustainability Minimum lending to startups Efficiency Ratio Organizational Capacity Organizational Capability Minimum lending to growing and internationalizing companies Outstanding BDC Financing Portfolio Distribution of lending by risk category (minimum percentage of lending to riskier segments of the market) BDC Consulting Revenue Employment creation (typically estimated by the companies) Return on Common Equity Minimum ratio between loans and guarantees Maximum lending to affiliates ECA objectives Efficiency cost ratio Capital adequacy ratio. Sources: BDC, BancoEstado, DBSA and Finnvera As shown in Table 2, governments/SFIs have different approaches for implementing the mandates. While the BDC mandate is well defined and the company is focused on the target sector, the performance indicators are more aligned with client satisfaction and efficiency. In DBSA's case, with a more broadly defined mandate, the performance indicators put more emphasis on aligning DBSA's business with a more specific mandate. Finnvera's corporate plan 14 South African privately owned commercial banks typically have a 10 percent return as a target, 13 imposes clear targets for the business areas that are in line with the government goals. BancoEstado's social objectives are diluted in the corporate plan and performance indicators. Consistency in performance indicators is an issue that requires further analysis. For example, BDC does not include job creation as an indicator, because it may contradict the objective of creating sustainable business in Canada. In the case of DBSA, the Treasury approves DBSA's corporate plan on an annual basis. The corporate plan includes also the corporate balanced scorecard, which has three components: development impact, sustainability and organizational capability. It is interesting to note that as a mechanism to focus the role of DBSA, development aspects have substantive weight in the overall performance measure of DBSA. As shown in Table 3, 50 percent of the development impact component is about the effectiveness of key development initiatives, which gives a clear signal of DBSA's priorities. Table 3. DBSA Corporate Balanced Scorecard 2008/09 Development Impact Sustainability Organizational Capability (56%) (24%) (20%) Development Fund Corporate credit rating Staff satisfaction with internal Performance (20%) (10%) knowledge management (25%) Non performing book debt Market perception of the bank's image Customer and partner as a percentage of total book in the development finance scene satisfaction (10%) debt (20%) (25%) Rand value of total Percentage of acceptable vacancy disbursement (10%) Sustainable earnings (20%) levels (25%) Co-funding ratio (10%) Cost to income ratio (20%) Share of total commitments to identified market segments (5%) Source DBSA While BDC's corporate plan is defined for a five-year period and requires the approval of the government, the balanced performance measures are used for the internal performance evaluation. The performance measures include the following indicators: client satisfaction; employment engagement; efficiency ratio;15 outstanding DBC financing portfolio (in dollars); BDC consulting revenue (in dollars); and return on common equity (ROE). These performance 15 Measured as the subordinate financing, operating and administrative expenses as a percentage of their net interest and other income. 14 measures are more focused on client satisfaction and the efficiency of BDC, but are consistent with a bank that has a well defined mandate. Finnvera's performance targets are defined by the Ministry of Employment and Economy in the annual corporate plan and include the following indicators: Minimum amount of guarantees and loans; minimum lending to startups; minimum lending to growing and internationalizing companies;16 distribution of lending by risk category (minimum percentage of lending to riskier segments of the market); employment creation (typically estimated by the companies); minimum ratio between loans and guarantees; maximum lending to affiliates; ECA objectives; efficiency cost ratio; and capital adequacy ratio. Finnvera's corporate goals provide clear guidance to the administration regarding the segments of the market on which the management needs to focus. The Ministry of Employment and Economy follows the accomplishment of these objectives on a regular basis. In the case of BancoEstado, the corporate plan highlights the bank as a universal (commercial) bank that provides all types of financial services to nationals. The government does not participate in the formulation or approval of the corporate plan. Performance indicators are more aligned with those of any other commercial bank, with only a hidden attention to the social objectives of the company. As shown in Table 4, the 2008 balanced scorecard included three components: financial results (30 %); client satisfaction (42%); and organizational capability (28 %). Only one social objective is included in the client satisfaction area, but refers to the liability management side of the bank (increasing the network of points of sales (caja vecina)). The balanced scorecard makes no references to other social objectives such as the development of products for micro and SME lending, and mortgages for low-income people. Table 4 BancoEstado Balanced Scorecard 2008 Financial Results Client Satisfaction Organizational Capacity (30%) (42%) (28%) Efficiency ratio Sales through Operational costs remote channels Average ROE of Multiple products to Internal control commercial banks clines Diversification of Index of client Engagement of sources of revenues satisfaction employees Access through points of sales 16 Internationalizing companies are companies that are looking for euro or world scale 15 2.3 Financing Policy Mandates In an ideal world, SFIs should be able to create a sustainable business within the boundaries of their mandate. With some exceptions, this has proven to be difficult and begs the question that if the business within the mandate is profitable, why it has not been exploited by the private sector. In many cases, SFIs have to rely on some other sources of financing, including cross subsidies, and direct support from the government. This section analyzes the effects of different sources of financing on the incentives of the management and board of directors of the SFIs. This section also discusses the importance of funding sources for the SFIs. a. Target sector is profitable The ideal way of financing the mandate is to develop a sustainable business model in the sector that the SFI is trying to support. This is a Pareto superior result, since the SFI is providing financial services to underserviced sectors, the management has the right incentives and there are no social costs attached. However, this is not easy to achieve, since it implies that the public sector is more efficient than the private sector in providing services to certain segments of the market. Some market failures may open opportunities for creating SFIs that are self sustainable with narrow business models, for example the public sector may have a lower discount premium than the private sector in projects with long maturity. Due to the lack of development of a yield curve, SFIs may be willing to take more interest rate risk and to be more efficient than the private sector in the financing of housing and infrastructure. b. Target sector generates operational profits Although some SFIs can create profitable business in sectors that are underserviced by the commercial banks, the revenues may not capable of financing the fixed costs of the project. In order to be able to run a profitable business, the financial services offered to target sectors require operating on the broader platform of the bank that includes distribution channels and risk management models. As the revenues of the business in the target sector might not be enough to cover these costs, SFIs may need to get involved in other areas of business to secure the necessary infrastructure that can make viable this business. Relying on broader platforms requires SFIs to diversify into areas different from the target sector. Creating a large SFI that gets involved in businesses that are more properly the province of the private sector, with the purpose of funding a broader platform to ensure the SFI's profitability might be costly from the social perspective. Finding the right balance between the benefit of providing services to underserviced sectors and intervening in private business is a major challenge for policymakers. Strong governance structures and mature risk management systems in the SFI are important preconditions that need to be in place before attempting to target sectors with operational profits through SFIs. Providing a proper diagnostic of the business model and identifying the synergies 16 that can be achieved within the platform of the SFI are essential for a successful involvement of an SFI in a specific sector of the economy. Development of a business model for the targeted sector may require sizable investments that are not easy to replicate with subsidies on transactions. In some cases, a sustainable business model for a specific sector of the economy might not emerge simply by subsidizing transactions from commercial banks because the development of a product require a number of studies and other investments whose costs are not easy to foresee by policymakers and are hard to develop in an experimental environment. Some may argue that once the financial technology for making profitable this particular business becomes available, it should be shared with (or sold to) the private sector. Although this seems a reasonable argument on competition grounds, it does not solve the problem of the intervention of the public sector in businesses that should be administered by private sector. The contrary argument is that SFIs may have a role in developing financial solutions for underserviced sectors, prior to having private sector involvement. c. Target sectors do not generate operational profits Governments might be interested in having SFIs that provide financing to sectors that are not profitable from the private perspective but have a high social rate of return. Although this is a valid option, the different alternatives impose externalities either on the market, the organization of the SFI, or in the incentives of the management. SFIs might finance these businesses from different sources: Cross subsidies from other business areas of the SFI. This model implies that the SFI needs to be involved in business areas that should be run by the private sector. As discussed above, this implies that greater involvement in business in which the private sector has comparative advantages. The effect is neutral if the SFI plays by the rules of the market, and competes with other banks in the financing of companies, but the situation is different if the bank operates with different criteria, for example offering below market rate lending to companies, which may end up crowding out participation of private enterprises and having a negative effect on financial development. Direct subsidies. The government may provide direct subsidies for lending in the targeted sectors. These subsidies can operate well, but in some cases might be insufficient for developing a sustainable business model for the SFI. These direct subsidies assume that SFIs have a business model that allows them to operate properly, which might not be the case. The main advantage of direct subsidies is that they could be offered to any banking institution, either private or state owned, and do not create distortions on the involvement of the SFI in what is essentially private business. Privately owned commercial banks can be slower in responding to these subsidies, as they may be unfamiliar with the entities to which the subsidies are directed or the entities are not part of their core business. The 17 provision of partial guarantees is an alternative way of giving direct support to lending operations in targeted sectors. Lines of credit with subsidized interest rates. Lines of credit are a powerful funding mechanism when no other alternatives are available. However, they create a critical distortion in the risk management structure of SFIs. SFIs that depend on lines of credit have the incentives to lend what they receive and to limit their operations by the amounts that the government is willing to offer depending on the budget. Lines of credit may create a dependence relationship between the government and the final destination of the funds, which creates room for political interference and mismanagement of the SFI portfolio. Erosion of the SFI's capital together with regulatory and supervisory forbearance. Some SFIs might be able to finance unprofitable projects with a softer regulatory and supervisory framework compared to other financial institutions. This mechanism allows for inefficiencies and political interference in the company. Poorly regulated SFIs are likely to be request periodic recapitalizations, which creates dependence on the owner and may compromise the independence of the board. SFIs should be subject to the same regulatory framework as other banks and therefore this approach should be avoided. Allowing SFIs to take deposits from the public. Due to lack of government funds, many governments allow SFIs to take deposits from the public. The decision about allowing SFIs to take deposits from the public should depend on the objectives of the SFIs and should be driven by the need to meet those objectives than the funding requirements of the bank. Only SFIs that have an objective related to the development of savings market should be allowed to take deposits from the public. It is necessary to take into consideration that deposit taking institutions require a more comprehensive supervisory framework than banks that only take wholesale funding. Most importantly, funding of an unprofitable business with deposits is unsustainable, and will require permanent recapitalization of the business. Consistent with the previous alternative, this mechanism should also be avoided. In the Canadian case, BDC is less risk averse than the private sector, and therefore willing to finance projects with higher risk. BDC has shown to be as efficient as the privately owned commercial banks, and has been able to create a profitable business from its target sector (Canadian SMEs). BDC has been maintaining a detailed database of its commercial relationship with Canadian SMEs since 1980. This has allowed BDC to understand better the risks inherent in these types of companies. In addition, BDC has developed sophisticated risk management system that has allowed them to price by risk and to run a sustainable business with SME. The management board of BancoEstado requires all business units of the bank to be operationally profitable, including its targeted sectors (e.g. micro and SME lending, and mortgage lending for low income people). However, these business units do not cover the fixed costs associated with the distribution channels, the back office, risk management systems, etc. 18 These business units can leverage from the existing platform of the bank, but are less likely to survive as niche banks. Since the clients of BancoEstado's targeted sectors are not within the risk tolerance of the privately owned commercial banks, this business is unlikely to be an attractive privatization prospect. As Finnvera's mandate includes both the development bank functions and export credit agency (ECA), the evaluation of its funding sources is more complicated. The ECA part has an automatic recapitalization component, which makes it difficult to evaluate the sustainability of the business model. Although Finnvera conducts rigorous assessments of all its lending and guarantees, the size of some projects (mostly from ECA part) is above the prudential standards. Thus, Finnvera has provided guarantees for amounts close to USD 1 billion, while the capital is about one third of that amount. Although Finnvera's capital adequacy ratios are above the required levels, the presence of the ECA business has consequently imposed some regulatory forbearance. From the bank sustainability perspective, it does not seem clear synergies exist between the ECA and the development bank, and we therefore we do not recommend this approach to other countries. In the case of DBSA, there are deliberate cross subsidies from the most profitable part of the bank (lending to large municipalities, financing of private projects and international operations) to the less profitable ones, such as lending to small municipalities.17 Financing of social projects is offered at a subsidized interest rate, while competitive projects are offered at market rates. All DBSA taxable profits go to DBSA Development Fund, and these resources are used to provide grants, deployment of technical expertise, training and identification of projects in poor municipalities. The funding sources of SFIs play an important role in the incentives for the board and management. Funding from the market by issuing bonds is a good practice as it avoids one of the powerful tools that governments have to try to influence the credit decisions of the SFI; increases legitimacy of the SFI with the privately owned commercial bank; and gives clear incentives to increase the efficiency of the SFI. Funding from the market is preferable when it does not have the explicit guarantee of the state. The four SFIs analyzed fund themselves from the market, but in some cases (e.g. BDC) the debt is with the full guarantee of the state. In other cases (e.g. Finnvera and DBSA), some of the bonds are issued with guarantee and other area not. It is important to note that in some countries the difference between one and the other can be substantial. The bonds issued by BancoEstado do not have explicit guarantees from the state. 17 The subsidies come in the form of lower interest rates 19 2.4 Measuring the Public Policy Performance of SFIs Since SFIs are established with a mandate to address a market failure, aside from evaluating the financial performance as other financial institutions, it is crucial to evaluate how well the SFIs fulfill this public policy mandate. In other words, it is important to evaluate the value added of SFIs compared with a case that they were not present. This feedback is useful not only to track the long-term impact of the SFI's activities on the market, but also to align future strategies with government policies. However, quantifying the contribution of an SFI in addressing these market failures can be a challenge. Most of SFIs analyzed have not incorporated this type of evaluation in their performance measures, there are only few institutions that are in the process of developing a number of indicators/proxies to measure their special contribution dictated by their mandate. We found that SFIs are using different benchmarks to evaluate their contribution. While DBSA or Finnvera plc compares their achievements against predetermined (but arbitrary) targets, BDC compares its results with industry benchmarks and compares its SME clients' performance with the performance of SMEs that received financing from other banks. While measuring policy performance against some predetermined target does not help to understand whether the SFI provides any value added, comparing performance with a sample of individuals / companies that did not receive the support of the SFI provides more insightful information about the contribution of the SFI. BDC has created a number of indicators to examine their value added in the financial market. These include "business creation," which compares the start-up authorizations BDC fostered compared with the market benchmarks; "business survival", which compares how many BDC supported start-ups survived two years and five years compared with the industry benchmark. Finally, BDC uses a "business growth" indicator to evaluate how BDC clients perceive their pace of business growth compared to the SME market in general. These indicators seem to provide more comprehensive conclusions about the SFIs contribution and impact than more typical indicators used by other banks, such as the number of new authorizations or volume for loans, grants or equity originated by the bank. In 2008, BDC commissioned Statistics Canada to conduct an assessment of the impact of BDC financing and consulting services on Canadian businesses. Statistics Canada measured the impact quantitatively by comparing BDC clients' performance to a group of non-BDC clients using four indicators: revenue growth, employment, business survival and profits growth between years 2000-2005. The results show that BDC clients achieved better results in terms of revenue growth, employment growth, and business survival than non-BDC clients. In addition, the report shows that the BDC clients' performance has further improved compared to non-BDC clients with every passing year after their decision to become BDC clients. Also, clients that used both BDC's financing and consulting services achieved even better results than the ones that used just 20 one type of service. Surprisingly, the results also indicated that the profit growth of BDC clients was lower than for non-BDC clients. However, this could be attributed to the fact that BDC provides longer-term financing and thus more stability, which may motivate BDC clients to increase their investment spending soon after securing BDC financing. One of the simplest ways of measuring the added value is evaluating the catalytic role in mobilizing and attracting funding from other financiers. DBSA and BDC uses a leverage ratio, which measures the ratio of the SFI's funding relative to other's funding, but this ratio is meaningful only if the SFI acts as the book runner of all the operations, otherwise it can be totally misleading. Other institutions are using some qualitative measures that evaluate the impact of the SFI on the overall institutional framework. For example, the International Financial Corporation (IFC) has developed a number of qualitative tools to evaluate the institution's accomplishment of the mandate, including the external impact on other variables, such as improved corporate governance, improved innovation or new regulation. This allows the IFC to identify the sequence of changes that occurred after their involvement, and determine, whether it was its involvement that lead to changes that, in turn, triggered an ex-post interest by private sector investors. While very important, the area of measuring policy performance still remains largely unchartered and will require further attention. SFIs need to be able to justify their existence not only for their financial result, but also for the achievement of their goals compared with a scenario that the SFI were not present. 3. Governance of SFIs This section analyzes the corporate governance aspects of SFIs that will protect the SFI against government intervention.18 Political intervention is one of the major threats for successfully functioning SFIs. This is typically a consequence of a lack of independence of the board of directors and senior management of the SFI and nontransparent communication between the SFI and its shareholders.19 While governments might be tempted to use the SFI to fulfill their short term political needs at the cost of affecting the financial sustainability of the SFI, they do not want to take responsibility for their involvement. The risk of political intervention can be reduced by ensuring that the representative of the shareholders is precisely defined. It is common to find SFIs where the entities and individuals 18 This section does not expect to serve as guidance for governance of SFIs. For an exhaustive revision of principles of corporate governance applicable to SFIs see Scott (2007). 19 Other state owned institutions face similar pressures, but the effect of these pressures on inadequate credit allocation can have more significant economic impact. 21 functioning as shareholders are scattered among a larger number of different government institutions, creating the conditions for multiple pressures on the board or management that result in credit misallocation or other inefficiencies. Political intervention can also be limited by ensuring a transparent and structured process for the nomination of board members. In order to ensure technical expertise, board members should be appointed in a transparent manner for a fixed period of time and they should be accountable for their actions in the same way as they are in public companies. The senior management of the SFI should be selected and be held accountable by the board of directors. The ownership policy, the nomination of board of directors, and the qualifications of the management are discussed in the following section. Defining the Ownership Policy The principal tasks associated with ownership include: participating in shareholders' meetings and voting upon matters within the authority of shareholders; electing or removing members of the board of directors; and monitoring the performance of the company. Scott (2007) suggests that in order to help ensure a consistent and comprehensive approach to government ownership, governments should develop and publicly communicate an ownership policy that itself will command the confidence of the markets. The policy should define the general terms and conditions which will apply to the government's investment and the manner in which government will exercise its ownership. The ownership policy should be based on a set of guiding principles that are associated with sound commercial practice and good corporate governance, as well as with competitive neutrality. Sound commercial practices and good corporate governance should be applied to decisions regarding the amount, structure, terms and conditions, and subsequent handling of the investment. In effect, the government should seek to mimic the practices of a private sector owner. A critical element in organizing the state ownership function is to make clear who is the responsible and accountable for representing the shareholder, and to define the shareholder representative's roles and accountabilities. In our four cases, there is a single Ministry responsible that acts as shareholder representative. Although we did not find explicit ownership policies in the four analyzed cases, various documents and regulations defined a number of principles applicable to ownership policies. The Ministry of Employment and Economy of Finland is the representative shareholder of Finnvera, and manages a portfolio of 21 state owned companies through the Corporate Steering Group inside the Ministry. The role of the Corporate Steering group at the Ministry of Employment and Economy is to steer of the various policy sectors of the Ministry and its administrative sector in order to reconcile their actions with the strategic goals set by the Government. The Corporate Steering group is active in coordinating the role of different agencies and SOEs, including Finnvera, to provide support to their clients. In many cases, the 22 coordination of support from multiple agencies can result into a successful financing of companies. The Corporate Steering Group is in charge of defining the annual objectives of Finnvera, and this group reports to the Undersecretary of State, who is also Chairman of the board of Finnvera. The decisions relating to the largest operations (typically from the ECA part of Finnvera) are not taken by the board, but by the Ministry of Employment and Economy or the Prime Minister directly. As mentioned before, Finnvera is a case of heavy involvement of the government in the company, but the risks are mitigated by cultural factors. The Minister of Industry of Canada (Industry Canada) is BDC's shareholder representative. The management of BDC is particularly interested in keeping the Minister of Industry informed about the activities of BDC. According to BDC's management, this allows them to avoid surprises and help BDC to run the bank smoothly. In this context, officials from the Ministry of Industry receive the same information as the members of the board of directors, and the management is willing to answer questions that the government may have. However, the management of the company does not take instructions from officials of the Ministry of Industry. The communications between the Minister of Industry and BDC take place directly between the Minister of Industry and the Chairman of the BDC. Policy direction from the Federal Government is generally provided via Priorities Letters from the Minister of Industry to the Chairman of BDC's Board of Directors. The priorities outlined in these letters influence the strategies and actions reflected in the Corporate Plan, which is ultimately recommended by the Board of Directors. The South African Treasury is the shareholder representative of DBSA and all the other development banks.20 The Public Finance Act sets out a comprehensive framework for transparency and prudent operations of SOEs, and the shareholder compact defines the state's ownership policy. The Treasury gets involved in the approval of the (annual) Corporate Plan and in focusing the mandate of the institution. For example, in the last few years, it has requested DBSA to focus its attention more on small municipalities. DBSA provides quarterly reports to the Ministry of Finance based on the plan, highlighting its achievements against the targets set in the corporate plan. The Chilean Ministry of Finance is the shareholder representative of BancoEstado. During good times, the Ministry of Finance plays a passive role as the shareholder of the bank. The discussions between the Bank and the Ministry of Finance are mostly about the distribution of dividends. During periods of recession, the Ministry of Finance has requested BancoEstado to pay particular attention to vulnerable sectors of the economy, but without interfering in the credit allocation process. On past occasions BancoEstado has also helped the government in providing liquidity to the banking sector. BancoEstado's management is proactive in finding financial solutions that may help with the objectives of the government. 20 The Department of Public enterprises is the representative shareholders of the other South African SOEs 23 Nominating the Members of the Board of Directors An appropriately constituted, qualified and empowered board of directors is an essential component for good governance. In most of the countries, board members of SFIs are appointed politically, which may result in lack of skills and mismanagement of the SFI. We analyzed cases where the appointment of board members has evolved to open and transparent mechanisms of selection of board members. While in BDC and DBSA the appointment of directors follow the best standards with high levels of transparency that promote independence and qualification for the positions, in BancoEstado and Finnvera, directors are appointed using political criteria. BDC and DBSA operate with a one board system, and have well developed system for the selection of board members. The Law establishes general fit and proper requirements for members of the board of directors. In both cases, a board committee prepares an assessment of the skill requirements of the board members, recommends skill requirements for the selection of new directors and assesses the capacities of the current board members. A shortlist is typically prepared by professional headhunters and the list of candidates is presented to the shareholder representative (the government). Although the government may decide not to select a candidate from the shortlist, and appoint an outsider, shareholders have typically accepted candidates from the shortlist.21 In both cases the Chairman of the board and the CEO are different persons. While no government officials participate on the board of directors of BDC, the Ministry of Provincial and Local and Government is represented on the DBSA board (but this Ministry does not have a direct role in the ownership function). In both cases the CEO is not appointed by the board of directors, which is an important shortcoming in the role and the responsibilities of the board. Finnvera and BancoEstado each have a two board system: a supervisory board and a management board (board of directors in the case of Finnvera). Despite being elected based on political criteria, supervisory boards of these two banks have limited responsibilities and have more an informational role. The supervisory board helps in informing the political world about the policy directions of the bank. Since the supervisory board has no responsibilities in the area of credit allocation, they are unable to exert any substantial political influence over the bank's decision. Regarding the Board of Directors, half of Finnvera's board is composed of government officials and the other half by representatives from the organizations in line with the company's industrial policy. Probably a board composition of this type in other countries runs the risk of appointing unqualified civil servants and sophisticated lobbyists from trade unions who would try to extract the maximum value of the bank for their individual constituencies, bringing the bank into bankruptcy in a short period of time. However, in the case of Finnvera, the board composition has brought highly educated public servants and responsible and qualified representatives from 21 When selecting outsider candidates, the government gets exposed to criticism for lack of transparency. 24 trade unions that run the company based on market criteria. Cultural elements and a good corporate law help to explain the effectiveness of the Finnvera approach. In any case, this is not a model that should be replicated in economies with weak governance systems and frail institutional design. A good practice found in the case of Finnvera is that its CEO is appointed by the board of directors. The management board of BancoEstado, including the CEO, is selected by the President of the Republic. Since the return to democracy in 1990, Chilean presidents have been careful to select individuals with technical capacity and political affinity with the government coalition. The model has worked well, and board positions have been rotated gradually with long overlaps among them. However, this procedure is subject to important weaknesses if the President changes the criteria for selecting the management board members.22 The management board acts as a triumvirate, as none of its members has individual responsibilities but the management board is responsible as a collective group. This system has forced agreements and coordination among the three members of the management board. A good corporate law and well defined responsibilities for board members have helped to align the interests of the board members with their mandate. In addition, BancoEstado is the only bank of the analyzed cases that is supervised as a commercial bank. Prudential supervision has been an important ally of the management board in promoting good practices in credit allocation.23 Ensuring Qualified Management A high level of qualification of the senior management is one of the common elements in the four cases analyzed. Management is essential for creating a risk management culture in the institution and capacity to operate under market criteria. The capacity for retaining qualified management is one of the elements that help to explain the good performance of these four cases. Compensation is typically based on market standards, although in some cases it could be slightly lower. In the case of BancoEstado, bringing talented management resources and removing the political appointees has been a long process, but it has helped to modernize the bank, allow it to become competitive, and build platforms for providing services to sectors underserviced by other banks. and Beyond... Some countries have designed additional safeguards to deal with political intervention. Clarifying the role of the owners, the board of directors, and the management leads to good governance which is essential for dealing with political interference, but sometimes it may not be 22 During the military regime in the 70s and 80s, the bank was used politically and its capital was severely eroded. 23 Many interviewed executives from BancoEstado praised the role of the Supervisor in helping to improve the standards of operations of BancoEstado. 25 enough. BDC, for example, as part of good governance practice, reports any undue pressure from politicians regarding its credit decisions to its Board of Directors.24 Since politicians seek to avoid making public that they have applied pressure on these institutions to fund projects that have ultimately proved to be unprofitable, this measure has dramatically deterred the political interference in BDC credit decisions. This safeguard has been able to successfully alleviate the pressure politicians may exercise on BDC successfully. Another example is the case of BancoEstado, which by Law is not allowed to lend to government agencies and state owned companies.25 This has limited the pressure on the Board of Directors to finance politically motivated unprofitable projects, and has promoted a market test for projects sponsored by the government. In addition, the Law does not authorize BancoEstado to give any type of subsidies. This has been an important tool to repel political pressures, and to encourage the Board to focus on efficiency. 4. Conclusions This paper discusses good practices for aligning the role of SFIs in solving market failures. The paper discusses the experiences of four leading SFIs, namely Canada's BDC, Chile's BancoEstado, Finland's Finnvera, and South Africa's DBSA, in defining, implementing, financing and evaluating their mandate, and in avoiding political interference through good corporate governance practices. The four SFIs analyzed have different checks and balances mechanisms for serving their target sectors and limiting political intervention in their credit decisions. However, they also share some common elements: firstly, they have an efficiency objective, which is measured on a regular basis by the shareholder. The four SFIs analyzed are required to be profitable on an annual basis and the shareholder or the board requires the accomplishment of some predetermined objectives. Secondly, they all have professional and qualified senior management supported by proper risk management systems. Thirdly, they do not depend on government support for their funding needs as, with the exception of BDC, they access the bond market on a regular basis.26 This last point is very important as it gives financial independence to the SFI, and removes the linkage between the shareholder and the allocation of funds. BDC is one of the strongest institutions from almost any point of view, with sound corporate governance and a sustainable mandate. The mandate is aimed at mitigating a market failure and is implemented in a way that complements the role of private financiers. The functions of the main stakeholders of the company are properly defined, and the nomination of board directors ensures that directors have the necessary independence and expertise. The government monitors 24 Undue pressure is defined as any behavior that goes beyond presenting a project for evaluation. 25 This concept also includes ministers and members of the Congress. 26 Although BDC get funding from the market through the Ministry of Finance, it does not depend on the government for passing them funds. 26 BDC regularly, and BDC management promotes owner activism to ensure that its actions are understood by the government. Lack of prudential supervision by a banking supervisor is a potential weakness of BDC, which is partially mitigated by a sound corporate governance structure and legal, regulatory and institutional foundations that resemble worldwide good practices. Despite the strong presence of government officials and lobbyists on the supervisory board and the board of directors, Finnvera is run in a prudent manner. This is mostly explained by cultural factors and sound laws that deal with the cases of potential conflicts of interest. The presence of qualified management and the trust on the board in the technical analysis and recommendations coming from the management are important elements that help to avoid political interference. Weaknesses of Finnvera include the lack of prudential supervision and regulatory forbearance. These weaknesses are partially offset by close monitoring by the shareholder against a number of short term objectives. The capacity of the shareholder to coordinate multiple mechanisms for servicing the final clients is also remarkable. Despite not having a clear mandate defined in the law, BancoEstado's management board has defined objectives, which are aligned with the social objectives of the government. BancoEstado has been able to develop sustainable lines of business in the financing of sectors of the economy that are not sufficiently well serviced by private financiers, including micro and SME lending. The risks of political interference are partially mitigated by rigorous prudential supervision by the Chilean banking supervisor, a lack of dependence on funding from the government, good corporate governance laws that define the responsibilities of board members, a structure of the management board that promotes consensus, a supervisory board with limited powers, and a number of clauses that do not allow the bank to give subsidies, and lending to government institutions. However, BancoEstado has important weaknesses in the areas of appointment of directors and in the description of its mandate. As happened before the 1990s, the lack of transparency in the designation of members of the management board and board of directors may bring the bank into an unsustainable equilibrium. In addition, since the mandate is not clearly defined, the bank may concentrate efforts into more profitable areas of business not related to the social objectives of the government, without a clear counterbalance from the owner. DBSA's sound corporate governance structure is one of its major strengths. This has facilitated the presence of an independent and qualified board of directors and a professional management. The shareholder also plays an important role in monitoring the short term objectives agreed in the annual corporate plan. The lack of financial dependence from government funding has serviced to strength the asset liability management functions of the bank. However, DBSA has some weaknesses in describing its mandate, and in finding a sustainable business model for the sectors in the target areas. DBSA lives with two conflicting objectives: financing infrastructure projects in middle income municipalities, and generating revenues to fund the operations of DBSA Development Fund. Although requiring a minimum return is a desirable objective inside a mandate, maximizing profits (with the purpose of funding DBSA Development Fund) may 27 intensify the investments of the bank in commercially oriented projects that are not necessarily consistent with social objectives. We noticed that little progress has been reached in measuring the public policy performance of SFIs. Despite the macro evidence that SFIs do not contribute to economic growth, there is room for making more narrow studies about the contribution of specific SFIs in solving a market failure. The results of BDC are encouraging and more research will be needed in this area. 28 References Beck, Thorsten and Ross Levine. 2002. Industry Growth and Capital Allocation: Does Having a Market- or Bank-Based System Matter? Journal of Financial Economics 64, 147-180. Caprio, Gerald, Jonathan Fietcher, Robert Litan, and Michael Pomerlano. 2004. The Future of State Owned Financial Institutions. Brookings Institution Press. Washington DC. Caprio, Gerald and Maria Soledad Martinez Peria. 2002. "Avoiding Disaster: Policies to Reduce the Risk of Banking Crisis." In Monetary Policy and Exchange Rate Regimes: Options for the Middle East, edited by Eliana Cardoso and Ahmed Galal. Pp 193-230. Cairo: Egyptian Center for Economic Studies De la Torre, Augusto, and Alain Ize. 2009. Regulatory Reform: integrating Paradigms. Draft. Office for the Chief economist for LAC and the Caribbean. The World Bank. Washington DC. De la Torre, Augusto, Juan Carlos Gozzi, and Sergio Schmukler. 2007. Innovative experiences in Access to Finance: Market Friendly Roles for the Visible Hand? Policy Research Working Paper No 4326. August. The World Bank. Washington DC Dinc, Serdar. 2005. Politicians and Banks: Political Influences on Government-Owned Banks in Emerging Markets. Journal of Financial Economics 77, pp 453-479 La Porta, Rafael, Florencio Lopez de Silanes, and Andrei Shleiffer .2002. "Government Ownership of Banks." Journal of Finance. Vol. LVII. No 1. February" Scott, David. 2007. "Strengthening the Governance and Performance of State-Owned Financial Institutions." Policy Research Working Paper 4321. The World Bank. Washington DC. 29