Partial Credit Guarantee Schemes to Promote Agricultural Finance1 Panos Varangis and Mazen Bouri Finance and Market Global Practice, World Bank Group Introduction This note has been prepared to provide practical advice and guidance to World Bank task teams, clients, and donor partners who are contemplating support to partial credit guarantee (PCG) schemes in order to promote access to finance for commercial agriculture. The note serves to highlight lessons learned from other PCGs, key features for the design of such schemes, and issues that should be considered during implementation. It does not include an exhaustive literature review, though it draws on evaluations already done. Ideally, this note would be reviewed together with the recently prepared note on matching grant schemes for agriculture; the two instruments complement one another, and they are being used jointly in a growing number of Bank-supported projects. Farmers, agribusiness small and medium enterprises (SMEs), and other SMEs in other sectors tend to be constrained in their access to credit by lack of collateral, credit history, and reliable financial accounts that would mitigate asymmetric information risks and enable financial institutions to better assess their creditworthiness. High interest rates on loans to farmers and SMEs reflect these conditions and inhibit credit demand, but even with high interest rates, supply of credit is greatly constrained for such borrowers, simply because financial institutions do not wish to take the risks. Moreover, most banks are not familiar with appraisal for small business credit, nor have they developed appropriate methodologies (such as credit scoring or cash flow–based appraisal) for assessing these types of loans; hence this market segment is not appealing to them. The negative impacts of credit constraints on the growth and economic performance of SMEs and farmers are well documented. PCGs aim to ease these constraints by absorbing part of the default risk of the borrower, thus providing financial institutions with a level of comfort that allows them to increase credit supply to credit-constrained firms and farmers. PCG schemes are widely used around the world. One study (FAO 2013) finds 2,250 PCGs in almost 100 countries, but the same study argues that specific PCGs are comparatively new for agricultural finance in emerging markets. PCGs often provide coverage for loans to SMEs more broadly, which can include agricultural SMEs among others, but there are also PCGs specific to agriculture (e.g., FAG in Colombia, ACGSF in Nigeria, and FIRA in Mexico2). There is also a review of the Thai PCG scheme by the Bank of Thailand, which points to improvements in access, lower interest rates, and lower collateral requirements. The note has benefitted from peer review comments by Simon Bell, Pietro Calice, and Michael Goldberg. Additional comments and suggestions are welcome and can be sent to the authors. 2 Fondo de Garantias Agropecuarias (FAG) in Colombia offers credit guarantees for loans to farmers and it is managed by FINAGRO, a public second-tier agricultural development bank. Trust Funds for Rural Development (FIRA) is a second tier development bank in Mexico that offers credit, guarantees and other services to the agriculture sector. The Agricultural Credit Guarantee Scheme Fund (CGSF) in Nigeria provides credit guarantees for bank loans to farmers and it is managed by the Central Bank. . 1 Advantages and Disadvantages of PCGs There has been some debate about the advantages and disadvantages of PCGs for agriculture specifically as well as for firms and SMEs in other sectors. The positive effects of PCGs relate to the financial additionality on the extensive margin, in that these guarantees induce banks to expand the risk frontier by providing loans to firms and farmers without access to loans (as opposed to simply subsidizing bank risk on borrowers who already have access to credit). PCGs also have financial additionality, meaning that guarantees improve access to better finance conditions, such as lower interest rates, larger loan sizes, longer loan terms, and less burdensome collateral requirements. The most common arguments against PCGs have to do with their possible negative effects on borrowers and banks, their financial sustainability, and the way they are managed. Among PCGs’ negative effects, the most obvious is their potential contribution to moral hazard, especially if they are not adequately priced: they might reduce the borrower’s incentive to repay the loan, particularly if the guarantee substitutes for collateral, and might reduce the lender’s incentive to properly assess credit and monitor the loan. The financial sustainability of PCGs is also a concern. First, the cost or premium PCGs charge to banks to access the guarantee often turns out to be too low to cover the real risks and administration costs. In addition, banks tend to access the guarantee for only the riskier loans, which creates issues of adverse selection. Finally, PCGs often are capitalized out of public funds (which often do not require returns), and they lack transparency and sound corporate governance structures. There can also be political interferences that can cause problems in the operations and financial viability of such PCGs. Various studies have shown some encouraging results on PCGs’ effect on extensive additionality—they seem to encourage banks to expand the risk frontier and provide loans to firms without access to credit—and on firm performance. Studies have also identified negative effects on moral hazard. But further empirical data are required to ascertain the effectiveness of PCGs. A recent World Bank study (Gutierrez, Hohberg, and Ortega 2016) finds that Mexican PCGs facilitate access to credit to riskier firms (smaller firms, firms with past defaults, first-time borrowers, and firms in poorer states). The study also finds that guarantees improve all loan conditions (interest rate, maturity, and loan amount). It does not find a negative effect of guarantees on defaults, and suggests that the limited moral hazard effects could be explained by the fact that guarantees do not substitute for collateral but complement it (guaranteed loans have higher collateral requirements than others; however, it is important to note that those loans would not have been provided absent the guarantee). But the study also finds low graduation rates from the PCG program—that is, PCG borrowers did not have easier access to nonguaranteed loans. In sum, the study finds positive effects of PCGs on financial additionality and no particular negative effects on moral hazard, but it also finds that firms do not seem to graduate from needing PCGs. The literature on PCGs indicates that they are not a panacea to solve the problems of access to finance by SMEs and farmers; but it also cautions against dismissing them as doomed to fail. To be most effective, PCGs should not be rolled out in isolation but should be part of a comprehensive approach to promote access to finance, one that addresses demand- and supply-side constraints as well the enabling environment. 2 Principles for PCGs Various study results have pointed to several key principles for good practices in structuring and operating PCG schemes. The FAO (2013) study argues that success of PCGs is aided by a strong and healthy financial sector with low levels of nonperforming loans (NPLs), transparent accounting, a good supervision and evaluation framework, and professional management free of political interference. More recently, a study by the World Bank and FIRST Initiative (2015, 10–11) has laid out 16 principles for public credit guarantee schemes for SMEs. These principles can be summarized as follows: • Establishment and operation of a public credit guarantee agency. If an independent legal entity is established to operate the PCGs, it should have adequate funding and a clear strategy, a proper legal and regulatory framework promoting public-private ownership, independent supervision, sound corporate governance, internal control (to safeguard the integrity of its operations and governance), and effective risk management systems (principles 1–8). • Targeting. PCGs should have clear and transparent eligibility criteria and qualifications for targeted beneficiaries (firms, SMEs, farmers, etc.), lenders (participating financial institutions, or PFIs), and credit instruments eligible for coverage. Beneficiaries should be firms or farmers likely to face financial constraints (first-time borrowers, younger firms, riskier firms or farmers, etc.). PCGs should focus on firms in certain sectors, such as agriculture, or in certain geographic regions that pose greater risk (principles 9–10). • Guarantee coverage ratio. Guarantees should be partial—less than 100 percent—in order to provide the right incentives for borrowers and lenders, and they should be designed to ensure prudential behavior by lenders (principle 11). Furthermore, a recent study shows that the coverage is usually in the range of 50–70 percent. Calice [2016] finds that the average coverage ratio of PCG schemes is 70 percent globally, 65 percent for upper-middle-income countries, and 66 percent for low-income countries. Guarantees should cover only loan principle to provide enough incentive for banks to monitor the borrowers. Coverage is most often pari-passu, although in some cases a first-loss guarantee mechanism is in place at the portfolio level. • Pricing. Guarantees should adapt risk-based pricing principles to cover the cost of risk and administration and to ensure longer-term financial viability of the PCG scheme (principle 12). • Claims process management. Following default, the triggering of claims and payment should be prompt, efficient, and transparent, but banks should also have sufficient incentive to implement vigorous loan collection procedures. Claims process management should take into consideration the country’s legal and regulatory framework and established loan recovery procedures for banks (principle 13). • Reporting, monitoring and evaluation, and impact. PCGs should be subject to rigorous financial reporting requirements and should disclose nonfinancial information related to their operations. In addition, their performance, outreach, and additionality should periodically be evaluated, and findings should be publicly disclosed (principles 14–16). 3 Key Design Features for PCGs Beyond the above principles, there are some additional issues that should be taken into account when considering the inclusion of PCG schemes. Although these design features apply to broader SME guarantee schemes, some attention on the specific context of agriculture needs to be considered. Institutional arrangements. Although having a specialized agency to operate PCG schemes is the preferred option, such an agency is not strictly necessary. Some countries operate PCG schemes through public development banks, central banks, or public-private partnership (PPP) development institutions. In some countries there is a specialized guarantee agency for agricultural loans. In the absence of a specialized agency, the financial institution (either public or PPP) that operates the PCG scheme should be one that understands credit evaluation and risk assessment, particularly for agricultural clients, and that has some track record in operating guarantees. The most prevalent form of ownership of the PCG agency is government, followed by a state-owned development finance institution, and then by a commercial financial institution (Calice 2016). Individual guarantees versus portfolio guarantees. PCGs can be assigned to each loan or can cover a portfolio of loans for the targeted group of beneficiaries. Individual-loan PCGs carry the risk that the borrowers may know (or find out) that their loan is covered, which may increase the moral hazard (risk of default). Knowing that specific loans are covered by guarantees may also encourage banks to seek the guarantee for riskier loans only and to be less careful in their loan assessment. This may be more of an issue if guarantees are not priced based on risk and their premium is below the true cost of risk. In contrast, portfolio-level PCGs are likely to have lower moral hazard for borrowers and lenders, as borrowers do not know whether their specific loan is covered. On the other hand, in portfolio guarantees, there is a risk that banks may include loans outside the target groups or loans they would have issued anyway. One benefit of individual guarantees is that they make it easier to track the beneficiaries and deliver any complementary technical assistance for the target group. According to the survey results reported in Calice (2016), about half of PCGs are for individual loan guarantees and only 10 percent are for portfolio guarantee; the rest are schemes that include both. So among the existing PCG schemes, individual-loan guarantees predominate. Technical assistance for beneficiaries and banks. In various projects, providing technical assistance to farmers and SMEs has significantly improved the quality and absorption of funds (use of guarantees). Technical assistance for beneficiaries aims to improve the quality of loan applications and make it easier for banks to assess the bankability of the proposed investments that need loans backed by guarantees. Technical assistance for banks increases their understanding of the agricultural sector and agricultural value chain, and helps them assess the systemic risks. The PCG addresses any residual risks and facilitate banks to provide loans. According to a UNIDO study (Green 2003), PCG systems that provide information and capacity building to banks about targeted borrowers (relating to cash flow, financial performance, business model, production technologies, and marketing) may enhance banks’ ability to recognize the target group as a new and potentially creditworthy borrower, while the lower risk of loan losses associated with the PCG encourages banks to provide the loan. Operational issues for PCGs. Although the general principles presented above are important, a number of specific details about operation of PCGs are also critical for successful execution. PCG operational manuals are needed to provide the level of practical detail and specificity required for carrying out the PCG schemes. Operational manuals should provide details on the following critical areas: 4 • Claims process management. Effective claims process management helps maintain a good balance in which the bank is incentivized to collect on the loans but also recognizes that payments need to be made promptly. In practical terms, achieving this balance is difficult. Banks often complain that PCG administrators require them to file cases in courts, pursue the collection of loans by any means, and carry out time-consuming verification of claims. These requirements discourage banks from using such guarantees. The operations manual should describe the process and time frame for actions that need to be followed, and should present specific steps that are consistent with the loan collection practices of the relevant country’s banking sector and with local laws and regulations that impact loan collection. In addition, the manual should specify the arrangements for post-claim settlements, when after the guarantee has been paid the banks manage to collect part of the loans. Calice (2016) reports that after a borrower misses a payment, the median length of time before the guarantee is triggered is 120 days, with significant regional variations in this value. • Pricing policies. It is important to ensure that guarantees are priced based on the riskiness of the client and activity they cover (risk-based pricing policies). Typical pricing ranges from 1.0 percent to 2.5 percent (usually of the outstanding amount of the loan per year), but pricing ultimately depends on the risk management systems of the banks and financial institutions, the riskiness of the targeted beneficiaries, and the activities or projects that need the guarantee. In the Middle East and North Africa (MENA), the average fee charged in guarantee schemes was 1.5 percent per year (Saadani, Arvai, and Rocha 2010). A diversified portfolio beyond agriculture could help in reducing risk and ultimately lowering pricing. • Targeting types of clients and eligible projects. These eligibility criteria need to be transparent, well formulated, and specific in order to ensure the PCGs create additionality. • Eligibility criteria for choosing participating PFIs. PFIs need to have adequate capital requirements and liquidity (as specified in the banking supervisor’s requirements) as well as appropriate credit risk systems, solid corporate and risk governance, low NPLs, and some experience in lending to specific sectors or subsectors. PFIs should also have the capacity to assess environmental and social risks relevant to agriculture. • Loan size limit. In most PCG schemes the loan size is capped to ensure that the beneficiaries are SMEs. In most schemes in emerging markets, this cap is below $1 million, often around $400,000 to $600,000. However, there are PCG schemes, particularly in developed countries, that cap loan sizes at well above $1 million and in some cases as high as $3 million (e.g., the United States, Republic of Korea). • Reporting. The manual should indicate what data and information should be collected and how often it will be reported. • Roles and responsibilities. The manual should provide details on the roles and responsibilities of PFIs and the PCG administrator/operator. • Capacity building. The manual should provide information on any capacity-building activities being provided to the PFIs or final target group. • Guarantee fund multiplier. PCGs are able to leverage funds and do not require $1 dollar of funds to guarantee a $1 of loan. The leverage ratio varies significantly depending on the country conditions, types and riskiness of the targeted beneficiaries, etc. For SME loans in developed countries, a leverage ratio of 1 to 10 ($1 of guarantee funds to back $10 of loans) or even higher may be possible. In emerging markets, leverage can be as low as 1 to 2.5, particularly for newer 5 types of targeted beneficiaries. A World Bank study for the MENA region (Saadani, Arvai, and Rocha 2010) found an average guarantee multiplier of 3.4. The average leverage ratio reported by Calice (2016) is 3.3. As a result of this leverage, the coverage provided by PCGs goes beyond the actual fund amount and expands the impact of this instrument. References and Background Sources Calice, Pietro. 2016. “Assessing Implementation of the Principles for Public Credit Guarantees for SMEs: A Global Survey.� Policy Research Working Paper 7753, World Bank, Washington, DC. FAO (Food and Agriculture Organization of the United Nations). 2013. Credit Guarantee Systems for Agriculture and Rural Enterprise Development. Rome: FAO. Green, A. 2003. “Credit Guarantee Schemes for Small Enterprises: An Effective Instrument to Promote Private Sector-Led Growth?� SME Technical Working Paper 10, UNIDO (United Nations Industrial Development Organization), Vienna. http://www.unido.org/fileadmin/import/18223_psdseries10.pdf. Gutierrez, Eva, Maike Hohberg, and Claudia Ruiz Ortega. Forthcoming. “How Effective are Mexican PCGs at Facilitating Access to Credit?� World Bank. Saadani, Youssef, Zsofia Arvai, and Roberto Rocha. 2010. “A Review of Credit Guarantee Schemes in the Middle East and North Africa Region.� World Bank. October. World Bank and FIRST Initiative. 2015. “Principles for Public Credit Guarantee Schemes for SMEs.� World Bank, Washington, DC. 6