Asymmetric information in securitization: An empirical assessment

Abstract

Asymmetric information in securitization deals is analyzed based on a unique dataset comprising a million mortgages, both securitized and not, and using a methodology, previously applied to insurance data, that looks at the correlation between risk transfer and default probability. The main finding is that, for given observable characteristics, securitized mortgages have a lower default probability than non-securitized ones. We show that this finding is consistent with banks caring about their reputation for not selling lemons.