2.2 The Problem of Asymmetric Information in Credit Markets
2.2.6 The Role of Collateral
contracts with lower interest rates and collateral requirements. Itmay, however, be difficult to accurately determine the risk of each individual borrower. Instead of writing individual- specific contracts, the lender may group borrowers into different risk classes offering a different loan contract per risk class (Hunte, 1993). Alternatively, the lender may employ a collection technology, which focuses on imposing penalties upon default to control for risk.
To employ a suitable collection technology implies effective contract monitoring. Devinney (1984), as cited by Navajas (1999a), argues that screening and collection mechanisms are substitute technologies and depend on the type of information available in the market.
In microfinance markets, the information problem is exacerbated by the absence of documented credit histories, standardised information and financial statements. Household and business expenditure are often not separated making it difficult to assess ability to repay as income may diverted from the business to the household or vice versa (Gonzalez-Vega, 1998). The absence of easy to interpret information makes it difficult and costly for MFOs to apply formal scoring techniques. MFOs have, therefore, relied more on contract design to create suitable incentives for borrowers to repay the loan, and on peer monitoring techniques to overcome information asymmetries (Conning, 1999; Rodriguez-Meza, 2000).
the right incentives, since effort is a disutility to borrowers and borrowers obtain greater expected utility from not having to bear any of the risk in the loan contract. Hence lenders will want to pass some of the responsibility of the project outcome to the borrower as an incentive to exert the desired amount of effort (Rodriguez-Meza, 2000).
Collateral is "an asset that upon liquidation is adequate to cover all or most of the lender's risk exposure including principal, accrued interest and collection costs" (Nagarajan and Meyer, 1995: 3). As an enforcement device, collateral secures loans against exogenous risks (poor business performance due to events uncontrollable by the borrower) by allowing the lender to liquidate the collateral in the event of loan default, reducing his default loss (Barro, 1976). Ex postcontractual risk is thus reduced. Ex ante moral hazard is reduced when threats of foreclosure discourage the borrower from engaging in moral hazardous activities (Bester, 1985). High-risk borrowers will, therefore, be required to offer more collateral than low-risk borrowers, while the loan amount is expected to increase and the interest rate to decrease as the value of the collateral increases and lender transaction costs in using the collateral decrease (Barro, 1976).
For collateral to be a useful incentive/enforcement device, it must be able to reduce the lender's default loss or make it costly for the borrower to default. This requires that the asset has well-established and transferable property rights, and a legal environment that facilitates loan contract enforcement such that the lender is able to foreclose and attach the asset.
Liquidation costs must be sufficiently low and asset marketability good to enable the lender to recover sufficient funds from liquidating the collateral to cover loan losses. The asset should also not be easily prone to loss of value due to collateral specific risks such as theft, pretended theft, damage by fire or accident and poor maintenance (Binswanger and Rosenzweig, 1986).
Although collateral can serve as an incentive device in a loan contract, Stiglitz and Weiss (1981) show that credit rationing still exists even with collateral. While for sufficiently low levels of collateral no adverse selection effect occurs, there exists a critical level of collateral above which low-risk borrowers drop out of the loan market. If this effect outweighs the incentive effect of collateral, then the credit market may still be characterised by non-price credit rationing in equilibrium.
However, if the lender is able to simultaneously vary the interest rate and collateral requirements, Bester (1985; 1987) shows that credit rationing may be eliminated since the lender can induce borrowers, by self-selection, to reveal their risk type. Hence a separating and not a pooling equilibrium will exist. Borrowers with a low probability of default will accept loan contracts with higher collateral requirements and lower interest rates, while high- risk borrowers will accept contracts with higher interest rates and lower levels of collateral.
Moral hazard will also be reduced since higher collateral encourages the selection of less risky projects.
For collateral to be an effective self-selection device, borrower wealth and collateralisation costs are important. Borrower collateralisation costs include the potential loss of collateral if the investment fails, costs incurred by the borrower in pledging collateral (group formation, legal costs) and foregone opportunities to use collateral to secure additional debt (Chan and Kanatas, 1985; Federet al., 1988). The ability of collateral to serve as a signalling device may be undermined if the marginal collateralisation costs of low-risk borrowers are higher than for high-risk borrowers, if low-risk borrowers have less wealth to offer as collateral than high- risk borrowers, and where no credible threat of foreclosure and the attachment of assets exists
(Bester, 1985; 1987; Besanko and Thakor, 1987; Chan and Kanatas, 1985). In addition, the asset holdings should be positively related to ability to repay so that low-risk borrowers can distinguish themselves from high-risk borrowers.
Bester (1994) argues that the potential for debt re-negotiation further undermines the ability of collateral to serve as a signalling device. Where this potential exists, lenders will require collateral to encourage borrowers to truthfully reveal the outcome of their project. In the presence of limited liability and the possibility of debt re-negotiation, high-risk borrowers may have an incentive to default if it is too costly for the lender to assume management of the project. Collateral reduces this incentive to default since the lender can take possession of the collateral to reduce loan losses. However, in this case both high- and low-risk borrowers have an incentive to pledge collateral reducing its value as an incentive device. Bester's (1994) model only holds true in the case of ex ante symmetric information.
While collateral can act as an incentive and signalling device, loan contracts are seldom fully collateralised. Excessive collateralisation costs, limited borrower wealth and poorly developed property rights in microfinance markets prevents the use of collateral to the point where the marginal disutility of the borrower's effort equals the marginal product of the investment (Besley, 1994; Gonzalez-Vega, 1998). This reduces the use in particular of formal collateral types such as land and chattel assets in microfinance markets. Costly legal processes in attaching and disposing of collateral also make it less attractive for MFOs to use collateral (Besley, 1994). In addition, loan contracts that require formal collateral types would also restrict MFOs to relatively limited, wealthy client market segments. Lenders may also want to assume some risk, since increased risk increases the expected return (Navajas, 1999a).
2.2.7 Loan Contract Monitoring
Monitoring of borrowers' actions is a possible way to ensure incentive alignment between borrower and lender, particularly in limited liability contracts where borrowers are unable to pledge sufficient collateral. Monitoring can take place during and/or at the end of the loan contract and may only be worthwhile if it lowers the potential for moral hazard (Conning, 1999). For monitoring to have any effect, it is important that the borrower's gains or losses are influenced by monitoring (Navajas, 1999a). Optimal contracts would involve borrowers repaying the lender according to some predetermined fee schedule, if the results of monitoring reveal that borrowers actions are appropriate, while borrowers receive a less preferred schedule (liquidation) should monitoring reveal that their actions are inappropriate, assuming perfect contract enforcement. Williamson (1986; 1987) shows that in credit markets where lenders have symmetric ex ante information about borrowers but asymmetric information about the outcome of the project, credit rationing will likely be the equilibrium outcome.
The quality of the information obtained by monitoring depends on the resources committed to monitoring, and the available monitoring technology. Thus different levels of direct monitoring may result where monitoring is expensive or where substitutes for monitoring are cheaper (Arrow, 1985; Pratt and Zeckhauser, 1985; Conning, 1999). For formal lenders in rural fmancial markets, direct monitoring may be extremely costly due to the geographic dispersion of rural borrowers, and may make lenders more reluctant to operate in these markets. Mechanisms to reduce monitoring costs are thus an important innovation that can contribute to improved financial service provision by MFOs.