Optimal design of securitization in a principal-agent relationship based on Bayesian inference for moral hazard

• Main Authors: Ezatollah Abbasian, Mohsen Ebrahimi, Elham Farzanegan
• Format: Article
• Language: fas
• Published: University of Tehran 2015-03-01
• Series: تحقیقات مالی
• Subjects: ibp stochastic process; moral hazard; mortgage-backed securities; nonparametric bayesian inference; optimal design of multi-task compensation scheme
• Online Access: https://jfr.ut.ac.ir/article_50708_d74f10f8eb9b81e5c7b0c1ffc555a436.pdf
• ISSN: 1024-8153; 2423-5377

In the securitization process, by selling the mortgage loans to risk-lover investors, originator can allocate the mortgage loans risk to them. In this case, originator may not have an incentive to screen out borrowers, resulting in the moral hazard problem. This paper, within a principal-agent framework, analyzes this agency problem. Investor, to reduce asymmetric information, uses compensation scheme for giving incentives to the originator and by using the Bayesʼ rule, deals with inferring various dimensions of undertaken effort, and incorporates her joint posterior beliefs of the pooled loans’ credit position data and inferred various dimensions of effort into the designing contract problem. The results indicate that the shape of optimal contract is a function of the information content of investors’ observations and inferred knowledge about efforts, suggesting that using additional information prevents originator’s opportunism, the originator more likely performs the obligated tasks when lending the loans to the applicants.