Matching for Credit: Implications for Econometric Analysis and Market Design

In many markets, agents form matches subject to rules. Take a bank that decides on rules to diversify jointly liable borrower groups with respect to their exposure to common income shocks. Such rules affect match outcomes by influencing both who matches with whom (direct effect) and who takes a loan (participation effect). I develop the key trade-off for conflicting predictions of extant models of moral hazard and adverse selection and estimate both effects separately. Group formation creates an endogeneity problem, but a new structural model is able to exploit the implications of market sorting to separate both effects. The identifying exclusion restriction is that the characteristics of all agents in a market affect who matches with whom, but the outcome of a matched group is determined only by its own members. This exogenous variation is similar to an instrumental variable. I find that diversification – i.e. preventing the grouping of borrowers with similar shock exposure – has an aggregate welfare effect equivalent to a 9 percent reduction in interest rates.