Collateral, two-dimensional moral hazard, and competition

This dissertation develops a theory on the use of costly collateral in a credit contract. Existing theories advocate the function of collateral as an incentive device to reduce moral hazard. However, they do not distinguish between the effects of collateral on moral hazard regarding \risk-taking" from those on moral hazard regarding \shirking". I study a model with two-dimensional moral hazard under two extreme market structures: perfect competition and monopoly. In the section \Risk-taking Moral Hazard" (Section 5), I isolate the effects of risk-taking from those of shirking. I also study adverse selection in which consumer has one of two types (good types and bad types). In the model, banks choose not to use collateral to mitigate risk-taking moral hazard. In a perfectly competitive market, there is a unique separating equilibrium with zero collateral for the bad-type borrower and positive collateral for the good-type. I show that the latter is caused by adverse selection. In the pure monopoly case, zero collateral is optimal for both types. In the section \Shirking Moral Hazard" (Section 6), I consider shirking moral hazard with a continuum of e ort levels, but no adverse selection. In this case, the second-best effort level depends on the trade-off between the marginal benefit of effort (in terms of increasing expected return of the agent's project) and the marginal cost (in terms of reducing limited-liability rents). When collateral is not very costly, it is possible that the good-type borrower exerts effort at a level that is higher than the rst-best, regardless of the market structure. In addition, a lower cost of collateral and/or more competition can improve Pareto efficiency of social welfare. In the section \Two-Dimensional Moral Hazard" (Section 7), I extend my results to the case in which both types of moral hazard exist. Regardless of the market structure, I show that collateral is not used to mitigate risk-taking but to encourage a more efficient effort level. But when the optimal collateral to induce e ort happens to be higher than the interest rate, collateral also reduces risk-taking as a side effect.