A Little Knowledge Is a Dangerous Thing: Model Specification, Data History, and CDO (Mis)Pricing

Abstract

Complex structured products especially collateralized debt obligations (CDOs) were at the center of the recent credit crisis. This paper explores two structural sources of CDO mispricing: modeling difficulty and data limitation. Comparing pricing outputs among models with different specifications, we find that using model with advanced default correlation assumptions and adequate data inputs can reduce the amount of model-implied AAA-rated CDO securities by 13% on average. This pricing difference has predictive power for the subsequent downgrading of AAA rated CDO tranches. These results indicate that CDO mispricing can be partly, but not fully, attributed to model misspecifications, as well as limited availability of historical data on CDO collateral assets.

Keywords: CDO; Model Specification; Data History; Default Correlation; Frailty

JEL Classification: G12; G13; E43; E44