PART C: CROSS-CUTTING ISSUES AND CONSTRAINTS

7. Addressing Constraints on Secured Lending

The analysis in Part A showed that an enabling environment in which uncertainties and transaction costs are kept to manageable levels is critical to the feasibility of term finance. Moreover, the difficulty of managing and coping with risks was pointed out, especially those of a systemic nature.

Elements of an enabling environment include: sound, stable macroeconomic policies; a legal and regulatory framework that ensures and promotes the health and efficiency of the rural financial system without introducing undue biases against rural financial intermediation; enabling, coherent agricultural- and financial-sector policies; and investments in rural infrastructure, such as roads, irrigation, communications and marketing facilities. The last element reduces the transaction costs of productive activities, as well as of the provision of support services such as input supply, extension, business development and financial services.

Though these elements are particularly important for term finance, they are preconditions for expanding rural finance and developing commercial agriculture in general. As such, they have already been discussed in other volumes of this series^[43].

This part focuses on four issues of particular relevance to enhancing the outreach and sustainability of term finance and that require action or support by governments and donors. Each of these issues is complex and would merit in-depth discussion that goes beyond the scope of this publication. However, due to their critical importance to term finance, some key issues will be briefly considered:

• constraints on secured lending (chapter 7);

• managing systemic risk (chapter 8);

• reducing asset/liability management risks (chapter 9).

Collateral issues have been largely bypassed by the recent expansion of microfinance, which has found innovative solutions to the absence of suitable collateral for small and short-term loans. Chapter 4 indicated that collateral substitutes could be used to secure a limited portfolio of medium-term loans, especially in a context of limited competition among formal lenders. However, it also pointed to the limitations of collateral substitutes in scaling up a term finance portfolio and in extending loan terms and reducing interest rates. In situations in which a proven demand for term loans exists and RFIs have successfully introduced them to farmers, the legal and institutional environment for secured lending may become a critical constraint on term finance.

7.1 Elements of Secured Lending

An enabling environment for secured lending refers to the creation, perfection and enforcement of security interests in rural assets:

• Creation of security interests refers to the legal base for including certain assets as collateral in the loan contract.

• Perfection of security interests refers to the filing of security interests in public registries.

• Enforcement of security interests refers to the foreclosure and sale of assets used as collateral in case of loan default.

As argued in chapter 4, lenders are generally not interested in foreclosing on collateral, not only because of the usual delays and transaction costs involved, but also because of the confrontational stance, militating against good relationships with borrowers. However, the ability to foreclose on and sell the collateral quickly in case of a serious default is not only important in limiting the losses of a lender, but also in controlling moral hazard problems and maintaining repayment discipline. The higher the transaction costs, delays and uncertainties regarding the creation, perfection and enforcement of security interests in assets, the less willing will a lender be to accept them as collateral. It may decide not to make the loan, or may require additional or varied types of collateral, such as urban real estate or cash. This may deprive viable clients of access to loans and leave profitable investment opportunities unfunded, leading to social and economic costs.

This section highlights some legal and institutional issues that limit secured lending in developing countries and then indicates areas for reform.

7.1.1 Constraints on the Creation of Security Interests

Important assets of farmers and rural SMEs cannot be used as collateral

Real estate is often the preferred form of collateral for larger, long-term credits, because these assets are more likely to maintain their value during the maturity of the loan (being less likely to deteriorate, devaluate or disappear). However, several problems can complicate the use of real estate in securing loans.

Farmers may not have formal rights (legally recognized and enforceable) to land or other fixed assets. This might be the case in customary land-tenure systems or if transfers of land rights through inheritance or sales have not been recorded properly. In other cases, land titles might exist, but legal restrictions regarding the transfer of property rights limit the collateral value of land. For example, in some countries land-reform laws restrict the sale of redistributed land^[44]. Homestead laws are designed to protect borrowers from losing their residences by protecting some property from being mortgaged or seized. Similar legal constraints can apply to tools and other means of production. Such provisions might be justified from the perspective of protecting the livelihoods of poor people. They do, however, restrict their choice to use their assets to secure term loans that could lift them out of poverty by enabling them to finance productivity-enhancing or risk-reducing investments. Inability to use their assets as collateral may force farmers to rely on other sources of loans, sources that may make smaller loans and may charge higher interest rates, but require little or no physical collateral. This problem may force capital-constrained investors into making less expensive but often less profitable investments, with shorter gestation periods. The distribution of economic activities is then tilted in favour of those who can offer the best collateral, preferably urban real estate.

In countries where land sales or rental markets are poorly developed, legal provisions for using movable assets as collateral (farm machinery and equipment, cars or livestock) may be of crucial importance^[45]. In many countries, however, such a legal base is either not in place or inappropriate legal provisions constrain its use in practice. For example, some countries require the specific identification and enumeration of property offered as collateral, i.e. identifying the specific animals in a herd, rather than accepting a floating security interest in the cattle, described only by their total monetary value, as allowed in developed countries^[46] (Fleisig and de la Peña, 2003).

The possibility of using movable assets as collateral is not only important to farmers, but also to equipment suppliers and agroprocessors, which are often an important source of seasonal finance in rural areas. Their access to term loans from banks is often restricted by the absence of legal provisions enabling the use of business assets such as inventory or machinery as collateral. This also limits their ability to provide term finance by selling assets on a deferred payment basis or by prefinancing the establishment of perennial crops (Fleisig, Aguilar and de la Peña, 1994).

Fragmented legal provisions for the creation of security interests

Several laws may govern secured transactions according to different types of assets or borrowers. This fragmentation may create problems in determining the priority of creditors against the same asset. For example, if lender A has the pledge on the future crop and priority under the Civil Code, while lender B has priority against the crop when harvested and placed in the warehouse under the Warehouse Law, who has priority when the farmer places the harvested, pledged crop in a warehouse? In this case, lenders might agree that the crop was valuable, but not accept it as collateral because of the possible priority conflicts^[47].

7.1.2 Constraints on the Perfection of Security Interests

Registration of security interests ensures that there is no senior claim on the asset and avoids the possibility of the asset being pledged to different lenders. It requires laws that clearly define priority and provide for registration of the security interest against the asset within appropriate registration systems. Usually, there are different registry systems for different types of assets, such as real estate, cars, ships and airplanes. Registries for other types of movable assets are often non-existent.

Various problems increase the transaction costs of using the legal registry system. Registries are often located in major cities far from investors and rural lenders. Many are not computerized and require manual searching. This creates problems for the lender, especially in the case of movable assets, which might be registered in another region, making detection of existing claims difficult. Fees are often high, because the lender must register the whole security agreement or a lengthy abstract, which is usually checked by registry staff for its legal correctness. In some cases, access to the registry is restricted. In some Latin American countries, for example, an official permission is required, which is a further source of delays and costs (Westley, 2003). This undercuts the very purpose of a filing system: making such data public. The accuracy of the data is sometimes questionable as well.

In sum, all of these constraints add to lender transaction costs^[48], which are then transferred to the borrowers, who face additional costs for registering their property. The high cost of creating a mortgage in most countries may make it uneconomical for small farmers to use their land titles to secure loans. Notice filing systems have been suggested as alternatives to conventional registry systems (Fleisig and de la Peña, 2003): here, the lender simply files a notice of the existence of the security interest. Information requirements are limited to names and domiciles of the parties and a description of the collateral. Given the limited amount and standardized form of the information, the database can be posted on the Internet. Notice filing systems are thus much easier and cheaper to maintain and are more user-friendly.

7.1.3 Constraints on the Enforcement of Claims

Enforcing claims against mortgaged property in the event of borrower default is often costly and time-consuming (see Box 22, p. 76). The greater the transaction costs, the less acceptable will be the collateral for securing small loans and the higher the collateral requirements. The repossession and sale of mortgaged assets usually requires court action that may take months or years. The sale of collateral is then subject to complex procedures, including an appraisal of the property, a court-administered auction and provisions for revaluation if the property cannot be sold at the appraised value (Westley, 2003).

In Tanzania the judicial process was found to take from three months to two years, and in India up to five years. In Bolivia the legal procedures for foreclosing on collateral take on average 269 days, though according to the law they should be completed within eight days. If the client refuses the order, the average increases to 670 days (Fleisig and de la Peña, 1994). This clearly affects the practical utility of collateral, and as a consequence lenders restrict collateral-based lending. Movable property tends to depreciate over time, so lengthy procedures further diminish its value as collateral. The enforcement period may exceed the economic life of some assets, such as standing crops and inventories, and there is the risk to the lender that they will simply be consumed, sold or made to disappear in the interval. Even when procedures are not inherently complicated, legal systems are often overburdened and inefficient, so that debt-recovery tribunals and courts function extremely slowly.

Some countries admit movable assets such as farm machinery and tools as collateral but exempt them from being foreclosed. Again, such a provision prevents farmers from using these assets to secure loans. This is especially problematic in situations in which land cannot be used as collateral.

However, there are other factors constraining the foreclosure and sale of land and other rural assets: In some countries, local politicians pressure banks not to sell mortgaged property, while community solidarity discourages people from purchasing foreclosed property. Markets for used machinery and other assets may be so thin that it is difficult to locate buyers or obtain the value.

7.1.4 Areas for Reform of the Legal and Institutional Framework

Collateral issues require careful identification, analysis, and reform as part of a longer-term reform agenda to strengthen the creation, perfection and enforcement of contracts for secured lending. Amending legislation usually involves a long process of consultation in order to find a compromise between different stakeholders and interest groups and to reconcile rural finance with other concerns. Moreover, the cost of reforms has to be weighed carefully against the benefits both in the short and long term, taking into account the institutional capacity of each country to implement legal provisions. A discussion of concrete measures for legal reforms and their sequencing would have to depart from the prevailing legal tradition and the existing legal system in individual countries and goes beyond the scope of this book. However, some possible areas of reform will be outlined.

Steps should be taken to broaden the range of assets that can be used as collateral. Depending on the current country situation, this might translate into diverse measures:

• In many countries, a legal framework and registry system for movable assets such as equipment, livestock, crops, inventory and receivables may need to be created or the existing one improved. This should include the possibility of creating floating security interests in livestock, inventory, etc.

• In countries without private land ownership, legal provisions may stipulate long-term use rights and leases for the securing of loans.

• Restrictions in land-reform laws and exempt property provisions on foreclosure and sale of rural assets should be carefully reviewed. The social benefits of protecting farmers must be weighed against the social costs of limited access to term loans. Instead of excluding the use of rural assets as collateral, governments may be better advised to enhance the availability of other risk-management instruments, such as the development of index-based insurance products and price-risk management tools.

Legal reforms should unify fragmented and sometimes ambiguous legal provisions and create a coherent legal framework and filing system for secured transactions, including leasing. Priority interests of different lenders on assets governed under different laws must be clearly defined.

Registration systems for filing security interests for real estate and movable assets should be reformed to enhance speed and convenience and reduce the cost of accessing information. This may involve the use of electronic data management systems where appropriate. In this case, however, additional costs for capacity-building have to be factored in. The feasibility of introducing notice filing systems might be explored.

Finally, legal provisions and administrative procedures should be reformed to ease contract enforcement and foreclosure on collateral. This could include non-judicial foreclosure, to save transaction costs and time, and the licensing of debt-collection agencies.

In countries where the basic institutional and cultural preconditions exist for rural land markets, the legal base for mortgage lending should be created. A mortgage law could stimulate the development of capital- market instruments, such as the securitization of mortgages on land and other types of asset-backed securities. This would open up additional sources of long-term funds for refinancing term loans for land development or the establishment of perennial crops.

As stated initially, legal reforms are likely to have the biggest impact on the supply of term finance in situations in which effective demand for term loans exists and RFIs have introduced them, but scaling up is constrained by missing or outdated titles, exempt property provisions or poor registry systems.

In view of the time horizons and complexity of legal and institutional reforms to expand secured lending, other measures could complement reform, including leasing. Chapter 5 highlighted legal- and tax-related discrimination against leasing, which merits attention and reform by policy- makers:

• Regulations limiting the types of institutions allowed to engage in leasing. Discrimination against financial leasing does not seem justified on risk grounds.

• Requirements that commercial banks open specialized leasing subsidiaries. This creates considerable additional cost, while the benefits are not always clear. The decision to create leasing subsidiaries should be left to FIs.

• Slow procedures for repossession of assets. In some countries, these reduce the comparative advantage of leasing over secured lending. Legal provisions may have to be reformed to ensure fast and non-bureaucratic repossession. In other cases, publicity campaigns and capacity-building of local courts and police may be needed.

• Unfavourable tax treatment. Tax treatment often makes leasing more costly than lending, especially when used as a financing instrument for small- and medium-sized farms in developing countries.

7.2 Guarantee Funds as a Risk-Sharing Mechanism

Guarantee funds are another alternative that might help FIs better manage idiosyncratic client risk. They might also provide a quicker response than legal and institutional reforms.

7.2.1 Principles

Policy-makers often advocate guarantee schemes to help lenders manage collateral problems. Credit guarantee programs are risk-sharing mechanisms intended to overcome the resistance of financial institutions to lending to targeted borrowers (individuals, households, farmers or small businesses). The objective is to stimulate lending to creditworthy borrowers that have feasible projects, but lack sufficient assets to offer as collateral. By sharing some of the lending risk, guarantee schemes are expected to leverage additional funds from the financial system ("induce additionality") because lenders make loans that otherwise would not have been made. If successful, they are more efficient in expanding lending than the provision of an equivalent amount of resources for refinance or loan revolving funds^[49]. Some programmes anticipate that, as a result of the guarantee experience, lenders will graduate to making these loans without guarantees once they learn that the targeted clients and/or investments are not as risky as originally perceived^[50].

A typical guarantee scheme links three agents or participants: a guarantor, a lender and a borrower. A guarantee agreement provides the lender with the right to call on the guarantee to recover loan losses. The design requires an incentive-based contract that encourages all three agents to act responsibly and prudently: additional loans are made, borrowers work diligently to repay them, and the costs and losses ultimately borne by the guarantor can be covered by fees and investment earnings, rendering the scheme sustainable. Ideally, a guarantee is only called after the borrower has done everything possible to repay and the lender has exhausted reasonable efforts to collect.

A successful guarantee programme may have favourable benefits for both borrowers and lenders. Besides receiving loans that they would not have otherwise, borrowers may benefit because interest rates may be lower, terms may be longer and collateral requirements may be reduced. Lenders may gain through reduced transaction costs and risks, larger loan volumes, and new clients, who become potential customers for larger loans in the future and for other products and services.

7.2.2 Empirical Evidence

Guarantee programmes for lending to small and medium enterprises (SMEs) have existed for several years in the U.S. and Europe. Many schemes have been introduced to expand agricultural lending in developing countries. More recently, guarantee schemes have emerged to support microfinance.

These examples are indicative of the chequered performance of guarantees. None of the European small- and medium-enterprise (SME) guarantee schemes are self-financing. Most in developing counties have failed or at best experienced limited success. Many failed because they were not financially sound and became moribund when they ran out of capital. Still others fell into disuse because the administrative arrangements were so expensive, complicated and time-consuming that lenders lost interest. Many schemes continue to depend on large subsidies, so they simply amount to credit subsidies dressed in different clothes.

7.2.3 The Way Forward: Better Practices in the Design of Guarantee Funds

Despite the problems cited above, there may still be circumstances in which a credit guarantee could be useful in leveraging agricultural term lending. This is especially the case in countries in which the absence of suitable collateral and legal and institutional constraints continue as major problems. Guarantee funds should not be used to circumvent underlying structural problems. They can, however, be important ancillary instruments to support the introduction of term loans. The key is to identify the indicators of success and in the design avoid the problems that have plagued previous efforts.

For the lender, the additional benefits expected from making guaranteed loans must exceed the additional transaction costs and risks. The costs include securing the guarantee, conducting the required level of client monitoring and loan collection, processing guarantee claims and absorbing the lender’s share of losses not guaranteed. The risks include those inherent in lending to the targeted borrowers and the investments they make, and the risk that the guarantor will be financially unable or refuse to pay the guarantee claims.

For the guarantee fund to be sustainable, investment earnings and the fee and commission income must be adequate to cover operating costs and guarantee claims. Sustainable funds will tend to be well-capitalized, earn high returns on invested capital reserves, receive high fee and commission income, and be designed and administered so they have low operating costs and pay few guarantee claims.

Several considerations are involved in reducing moral hazard:

• Borrowers need to have something at risk and should be required to offer some collateral even if it is not normally required for a loan.

• Borrowers that default must be denied further loans from all lending institutions.

• Lenders should have at least 30-40 percent of the loan at risk, while guarantees of less than 50 percent may significantly reduce the interest of banks to participate in the scheme.

• Loan payments should be allocated in proportion to the risk shared and not credited fully to the lender’s risk portion.

• Borrowers might be classified by risk level and the structure of fees and guarantee coverage set at higher levels for the riskiest borrowers.

An important conclusion is that SME guarantees cannot be imposed on a faulty financial system, unsound and inefficient financing institutions, and a general culture and legal system that condone non-repayment of debts. Lenders must be interested in learning to work with the sector and willing to commit themselves to developing the expertise to make good term loans. Without such a commitment, a loan guarantee scheme alone will not contribute much additionality nor develop a system of sustainable term lending. Moreover, no guarantee fund can solve fundamental structural problems if lending to agriculture is inherently too unprofitable and risky (Vigano, 2002).

Guarantee funds might be most effective if combined with the provision of technical assistance to the lender. If the designers of guarantees are correct in arguing that it is possible to identify borrowers and investments that are less risky than perceived by lenders, then it is necessary to train lenders to use appropriate client screening and lending technology. The guarantee then provides additional security to the lender in the short-term but becomes unnecessary in the long-term, once the improved lending technology is in place. Guarantees could therefore be structured as a temporary risk-sharing instrument to cover some of the high initial risk of institutions introducing term finance products. This may create further incentives for RFIs to put their own resources at risk and invest in staff training, product development and adjustment of operational procedures.

7.3 Credit Bureaus

Specialized institutions that collect and disseminate information on borrowers’ credit histories, credit bureaus are another potential innovation for improving access to credit. As a source of information, they can produce multiple benefits by reducing lender transaction costs and risks, promoting greater transparency on the financial transactions and obligations of borrowers, inducing greater competition among lenders for good clients, and creating incentives for timely loan repayment, as borrowers come to appreciate the value of a good credit history. Credit bureaus may take on greater importance in countries where it is politically difficult to change laws and regulations to make it easier and cheaper to use collateral to secure loans^[51]. A number of start-up bureaus have sprung up in West Africa since 2000, but it is too early to see results.

Many developing countries, especially in Latin America, have developed private and public credit bureaus. They collect information on credit histories from affiliated financial institutions, assemble it into a standardized database, and for a fee provide their affiliates access to individual credit histories. This information is used by affiliates to screen clients and make loan decisions. In some countries, all regulated financial institutions must furnish data to public credit bureaus. Not all financial institutions become affiliates, however. For example, MFIs may not be members either because only regulated institutions are allowed to participate, or they find it too expensive for the benefits received, or the size of their transactions does not reach the minimum accepted for registration. Moreover, some institutions prefer not to participate because they have expended great effort accumulating information about their own clients and fear they may lose the best ones if the competition gets access to this private information.

Credit bureaus might be an effective way of reducing the costs and risks of term lending, but it is unlikely that they will be relevant in the near future, considering the nature of rural finance in most countries. They are most useful where there are many competing lenders, serving many clients, so that the total number of transactions is large and it is difficult for financial institutions to directly exchange information. Consumer credit typically represents this type of situation. Microfinance may be entering this stage of development in some urban areas, and even in some rural ones, as in the case of Bolivia. However, the rural finance situation in which term finance might be introduced does not always display these features. There is often only one or a few institutions involved, and usually they do not compete with one another in specific localities. Moreover, potential clients normally engage in only a few financial transactions, and they tend to use informal or microfinance sources that would not participate in credit bureaus. Term lenders in this situation can easily obtain information directly from the few relevant financial sources in order to compile credit histories on potential clients. Only at later stages of development will urban-based credit bureaus begin to expand into rural areas and serve important functions for rural term lending.

8. Managing Systemic Risk

The analysis in Part A pointed out that one of the core constraints on term finance is the inability of investors and financial institutions to properly manage and cope with systemic risk. The measures discussed previously - an improved legal and institutional framework for secured lending, credit bureaus and risk-sharing mechanisms such as guarantee funds - are more suitable to managing idiosyncratic risk. The absence of proper instruments for managing systemic risk may force investors and RFIs into costly risk-avoidance strategies, foregoing opportunities for investment and growth. This underscores the importance of designing complementary risk-management instruments that address systemic risk.

8.1 Policy Options

Governments may use various policy measures and instruments to reduce the level of systemic risk and strengthen the capacity of investors and financial institutions to manage and cope with its impact. The level of uncertainty can be reduced through stable macroeconomic and agricultural- and financial-sector policies and by abstaining from ad hoc political interventions in financial markets. This allows investors and FIs to concentrate on managing climate- and market-related risks. Governments can further support the risk-management strategies of farmers through the following measures:

• investing in irrigation and flood protection. The rehabilitation of irrigation systems may have important secondary effects on demand and supply for term finance. However, capital investments in infrastructure are only sustainable if accountable, efficient management structures are created and farmers participate both in management and in operational and maintenance costs;

• researching crop varieties that are resistant to pests or more drought tolerant;

• strengthening the input supply chain to enhance access of farmers to drought- and pest-resistant seed and planting material and to agrochemicals;

• promoting price-risk management instruments such as hedging, forward contracts and put options in order to protect against short-term price fluctuations; and

• introducing innovative approaches to crop and livestock insurance.

The suitability of these diverse strategies depends on the severity and frequency of risk. For example, improved varieties and agricultural practices or investment in water-management facilities may to a certain degree enable farmers to cope with frequent but less severe climatic risks or pests. Crop insurance may cover systemic risks that are less frequent but cause higher losses. Major external shocks such as drought or the temporary collapse of key export-commodity prices cannot be insured against at a reasonable cost and may require temporary government or donor assistance to mitigate the consequences on producers, financial institutions and the national economy.

Though necessary in extreme cases, disaster assistance should be used with care in order to not crowd out other risk-management strategies such as insurance. If free disaster assistance is periodically available, farmers are protected from the costs of coping with systemic risk. In the long-term, this may result in a distortion of their investment decisions into high-risk activities. Thus, if free aid is frequently available, disasters may become self-perpetuating.

Financial institutions respond to systemic risk primarily by diversifying their portfolio into different sectors and regions. The trade-off between portfolio diversification and the need to invest in a specific financing technology, thereby acquiring local knowledge to manage credit risks, has already been mentioned in chapters 1 and 3. This applies especially to RFIs of limited size. Access to refinance facilities in the case of major external shocks may help those RFIs with a viable term finance technology cope better with systemic risk. If idiosyncratic risks are manageable, this would allow them to expand their term finance portfolio. Such refinance facilities could be made available through second-tier or apex institutions. However, refinance should be restricted to extreme situations and care should be taken that only well-managed, financially sound RFIs qualify for such facilities. This will discourage laxity in loan appraisal and bail-out mentalities among lenders. Debt forgiveness programmes may present the same pitfalls as disaster assistance, as well as undermining repayment discipline and careful scrutiny during loan appraisal.

8.2 Innovations in Agricultural Insurance

Innovative approaches to using index-based crop and livestock insurance might warrant attention and support by governments and donors.

Problems of conventional crop-insurance programmes

Agricultural insurance, and particularly crop insurance, has had a chequered history in developing countries. Along with lenders, agricultural insurers face moral hazard risks, adverse selection and high transaction costs, which are difficult to resolve under developing-country conditions. Some of the main problems of crop insurance are:

• Insurance of a large number of perils makes verification of claims extremely difficult, increases problems of moral hazard and leads to an increase in the number of claims.

• Verifying claims in scattered smallholdings and the processing of large numbers of claims are time consuming and costly.

• Public crop-insurance programmes often come under political pressure similar to that experienced by agricultural development banks, i.e. to make high disbursements in election years.

These factors lead to high costs that are seldom covered by premium payments. Thus public crop-insurance programmes in both developed and developing countries have been highly dependant on public subsidies. In most cases, the benefits to society have not been clear enough to justify these amounts of public resources (Skees, et al., 1999).

New approaches: area- and index-based crop insurance

Novel ways of overcoming these difficulties through area- and indexbased insurance products have recently been suggested. In these, the trigger for indemnity payments and the assessment of loss are done not on an individual basis (which is prohibitively expensive for small-scale farms) but rather on either area yields or some objective weather event such as temperature or rainfall. Provided such data are secured against fraudulent adjustment, the cost of setting up and operating the trigger is very low. Moreover, moral hazard risks are eliminated so that deductibles - commonly used in traditional crop-insurance programmes to combat such risks - are not needed. Naturally, only some perils lend themselves to this type of approach. Drought is a prime candidate.

The effectiveness of an index-based insurance contract depends on how positively farm yield losses correlate with the underlying area- or weather-based index. Due to the incidence of microclimates or the different factors influencing area yields, the actual extent of a loss may vary from the trigger. The so-called base risk describes situations in which an insured person suffers a loss but does not receive a payment or vice versa. To calculate premiums adequately, index contracts require reliable data at the local level in order to determine homogeneous areas, or correlations between losses and weather events. Thus the key is to find areas that are homogeneously affected by an external event such as rainfall. Index-based insurance policies could be sold either to individual farmers or to RFIs. The latter possibility would to a certain extent avoid base risks affecting individual policy holders, and it would protect the lender against losses from default due to drought (Skees, 2003).

To date, there are few examples of index-based insurance systems in developing countries, and most experiences are still at too early a stage to draw strong conclusions. However, given their potential to allow farmers and financial institutions to manage climatic risks, governments and donors might consider supporting their introduction on a pilot basis in order to test their feasibility under specific local conditions^[52]. This would require technical assistance and seed funding. Technical assistance could include capacity-building among interested stakeholders - farmer organizations, local insurance providers, RFIs and government departments - regarding the concept of and institutional approaches to providing index-based insurance. Local insurance providers could be trained in identifying suitable indices, using information technology to monitor and analyse data in order to design contracts attractive to farmers, while ensuring the financial viability of the providers. Seed funding might serve a reinsurance function during the start-up period until the schemes have reached maturity. Subsequently, governments and donors might facilitate access to reinsurance or global weather-risk markets.

9. Reducing Asset/Liability Management Risk

Section 3.8 concluded that FIs should fund their term asset portfolios through long-term funds at stable costs in order to reduce asset/liability mismatches and the associated risks. Equity, subordinate loans and long-term borrowing in the national currency are the most suitable funding sources. However, these funds might be in short supply in developing countries with poorly developed capital markets. Also, certain types of funds are only available to regulated FIs or might be more suitable for larger FIs, as was discussed previously.

Three situations can be distinguished in which governments and donors might have a supporting role in mobilizing funds for term finance:

• the introduction of term finance by FIs with a limited track record in the financial system;

• refinance facilities to accredited FIs for providing long-term loans with grace periods; and

• refinance facilities to smaller FIs with viable term finance portfolios, but limited possibilities for diversifying their funding sources.

9.1 Supporting the Introduction of Term Finance

Access to long-term funding sources on concessionary terms^[53] is particularly important for young FIs with a limited track record in the financial system. At this stage, while systems and staff capacity are still being developed and a reputation built, special assistance is almost certainly required. The availability of long-term funds at stable costs avoids interest-rate and liquidity risks and allows financial institutions to focus on developing a sound lending technology. This also applies to FIs with a successful track record in short-term lending that are interested in introducing medium-term loans or leasing. Once the lending risks are manageable and a suitable financing technology is in place, funds from donors can be provided on more commercial terms and complemented with other instruments such as borrowing and equity from commercial sources.

Access to equity, subordinate loans or long-term loans on concessionary terms doesn’t only reduce ALM risks. It may also partly compensate the RFI for higher initial unit costs per term loan or lease and avoid transferring these entirely to the client. FIs with a small term finance portfolio and limited experience, who still have to optimize their financing technologies, face higher costs and have to make higher loss provisions than experienced financiers. This translates into high interest rates or appraisal fees, which may reduce effective demand and leave many potentially viable investments unfunded. Moreover, adverse selection might become a problem in situations in which a financier has not yet fully developed the skills and instruments to properly appraise the financial viability of term loan or lease applications.

Through concessionary funding for the introduction of term finance products, governments and donors share the institutional learning costs. Such direct financial support to individual FIs might be indicated in situations in which term finance is not available to farmers and other SMEs in rural areas, despite the existence of effective demand. In view of the public good elements of the introduction of financial innovations and broadening of the financial infrastructure, this support could be justified from the perspectives of financial-system development and broader rural development. The successful introduction of innovative term finance products not only benefits individual FIs and their clients, but also has important spillover effects. On the demand side, financing of term investments may trigger intensification of production, value adding and the associated benefits of employment creation. On the supply side, the successful introduction of term loans or leasing by one RFI may prove the viability of such finance. Successful approaches might then be replicated by other FIs in the same country or elsewhere.

The case studies underline the importance of concessionary funds and equity in the early stages of an institution’s operational life, for example BASIX in India, which was originally capitalized by soft loans from donors. Similarly, the leasing programme of CECAM in Madagascar has depended heavily on long-term, cheap resources provided by the Government and the European Development Fund. The Bolivian MFIs received equity, subordinate loans or long-term loans from donors and international development finance institutions during their institutional start-up phase, and eventually for the introduction of medium-term loans and leasing. Borrowing on commercial terms has only been used after the start-up period.

Experience with credit lines in the past highlights the importance of using and designing such instruments carefully. In order to minimize distortion, concessionary funding should honour the following principles:

• Graduality. Concessionary funds should be provided only on the basis of a genuine, proven demand. It is important that the management of FIs should never be placed in a position in which certain lending targets have to be met. Term finance requires skill and experience in assessing and managing risk and costs. If these are not sufficiently developed, then the push to lend will be paramount, and loss of quality in the term loan asset portfolio will be the inevitable result. It is also important that the RFI puts some of its own resources at risk.

• Performance. Portfolio quality and performance should be closely monitored, and technical assistance may be necessary. Financial assistance should only be provided to the extent compatible with effective demand and the institutional capacity of the RFI.

Complementary measures. When a lender has established a viable financing technology and a reliable pool of borrowers, support may be required to access other sources of funds. Technical assistance may be needed to strengthen the ALM skills of the RFI. Guarantee funds may offer another possibility for supporting small RFIs with a limited track record in accessing commercial funding sources.

9.2 Funding Long-Term Loans

Investments with long gestation or amortization periods, such as planting of certain tree crops, farm restructuring or processing facilities, usually require long-term loans, possibly with grace periods. While experienced FIs with good ALM skills may fund a medium-term finance portfolio out of short- or medium-term liabilities, the provision of long-term loans requires long-term funding sources to avoid severe liquidity risk. These funds are seldom available in financial markets in developing countries.

Table 7
Change in BAAC’s sources of funds structure, 1967-2001

Source: BAAC

In Bolivia, for example, where several financial institutions have developed the capacity to provide medium-term loans with a maturity of up to five years, the provision of long-term loans is constrained by the lack of long-term refinance facilities^[54]. Even in many western countries, commercial banks provide few long-term loans due to their heavy reliance on short-term funds. Long-term finance is provided mainly by specialized institutions such as development banks and mortgage finance institutions. As long as the conditions for the development of mortgage-based securitization and other capital-market instruments do not exist, access to long-term refinance facilities might be crucial to enabling retail financial institutions to make such loans (particularly those with grace periods).

9.3 Funding Small, Successful Term Finance Providers

Access to refinance facilities with longer maturities might be important to smaller rural financial institutions with a proven term lending technology, but limited possibilities for diversifying their liability structure. This may be due to legal and regulatory constraints or to the high cost of long-term capital-market instruments such as bonds, which only amortize over a large portfolio. Using short-term funds such as commercial borrowing or deposits might still involve too high a risk to finance a significant term finance portfolio.

If there is a critical mass of retail lenders able and willing to expand rural term finance, the feasibility of second-tier refinance facilities such as apexes or wholesale banks might be explored. These can provide credit lines for term loans or leasing to eligible retail intermediaries, which would have to carry the lending risk. They may also provide training and capacity-building in ALM and other fields of banking. Such wholesale institutions could provide retail lenders with an opportunity to refinance a bad portfolio in the case of temporary, major external shock affecting a number of their clients (for example, drought or floods). However, this has to be assessed carefully case-by-case in order to ensure careful lending decisions by the retail institutions and discourage bail-out mentalities.

If long-term funds are provided to expand the existing term finance portfolios of retail lenders, they should be priced at market interest rates, above the interest rate of long-term deposits. The cost of funds should be adjusted periodically to changes in market interest rates. Moreover, it is important to ensure that a significant part of the credit risk is carried by the financial institution. The concessionary element would refer to the maturity, but not to the cost of funds.

In the long run, the development of capital markets should be stimulated, including mortgage-based securitization. Apart from sound and stable macroeconomic policies, this would require an enabling legal and regulatory framework, as well as effective supervisory institutions.