In this article we consider the problem of allocating microcredit among a set of networked cooperating agents by a social planner. Network structure, risk aversion and risk sharing of the borrowers have been analyzed in the context of microfinance. We propose a partial liability model, where the borrowers need to take responsibility in case of defaults by their immediate connected neighbours, by committing guarantees as social collateral ex ante to the social planner. We have developed a utility-based allocation model that is Pareto optimal and maximizes a social welfare function. The methodology computes the guarantees and the allocations for the borrowers. We show that the allocation formula for arbitrarily connected networks captures the notion of dominance over neighbours which are one hop away. We also show how the synergy between connectivity, risk aversion and guarantees impacts allocation. We develop the properties of the rule and illustrate them with suitable examples.

No abstract

No abstract

No abstract

The present paper is an updated and revised version of an internal unpublished technical paper developed as part of the research activity at the organization. The authors would like to thank the anonymous referees for helpful comments on an earlier draft.