Evaluation of Individual and Group Lending under Asymmetric information

The paper attempts to find the socially best loan contract by comparing exante welfare, interest and default rates of individual and group lending. We introduce a general framework which allows auditing policies and interest rates to be simultaneously determined by maximising the social welfare. Both variables vary with the types of risk considered: independently identically distributed and positively correlated risk. An individual project outcome is private information of its owner, but reported outcomes can be audited at a cost which then publicly reveals the true project outcome. We find that incentive compatibility in a group loan context is delicate: the conditions for truth telling vary with the borrowers’ perception of the overall solvency of the group. In addition, group loans are often made to local groups who have established local networks. This may mean that the group has cheaper policing of truthtelling, but also that the risks on projects within the group are likely to be correlated. To explore this, we numerically solve for the optimal contracts with varying audit cost differences and correlation, using a betabinomial distribution. We find that with an audit cost advantage, small group loans (typically to two borrowers) dominate individual loans even with correlation. But if audit costs are identical, the individual loan dominates. In the larger the group, the higher the audit probability is required to ensure truthtelling. Our finding provides an argument for why the number of borrowers should be limited to 2-5.