“A Little Knowledge Is A Dangerous Thing: Model Specification, Data History, and CDO (Mis)Pricing” by Sarah Qian Wang
Dragon Yongjun Tang
The University of Hong Kong
Sarah Qian Wang
The revaluation of collateralized debt obligations (CDOs) plays a significant role in the ongoing 2007-2009 credit crisis. Starting in August 2007, a large amount of initially AAA rated CDO securities are substantially downgraded, some directly to junk grade. This paper explores two structural sources of CDO mispricing: modeling difficulty and data limitation. Simulating the frailty correlated default model of Duffie, Eckner, Horel, and Saita (2008), we show that CDO mis-pricing can be partly attributed to model misspecifications, as well as limited availability of historical data on CDO collateral assets. This simulation result is consistent with empirical evidence on historical performance of a sample of 279 CDOs. The frailty model estimated with adequate historical data would have reduced the amount of AAA rated CDO securities by 18% on average. The frailty model has predictive power for the subsequent downgrading of AAA rated CDO tranches. Our study addresses practical issues on financial innovations and provides guidance for corresponding risk management.