The expansion of credit default swaps insuring mortgage-backed securities backed by subprime loans is shown to have a positive effect on the boost in subprime mortgage defaults. The authors provide the first empirical investigation of this relationship as it pertains to defaults during the 2007–08 financial crisis.
Many blame Wall Street (i.e., US financial markets) for its part in the subprime mortgage crisis that led to the Great Recession, claiming it overlooked the quality of subprime loans that were securitized, which incentivized originators to lend to riskier pools of borrowers to satisfy the appetite for mortgage-backed securities (MBSs). The authors reveal a more nuanced picture; their empirical analysis shows that the expansion of credit default swaps (CDSs) fueled greater demand for subprime mortgages that eventually defaulted because CDSs provided a way for market participants, specifically originators and securitizers, to limit their risk exposure to these riskier loans. Thus, the expansion of CDSs further propelled subprime mortgage defaults.
How Is This Research Useful to Practitioners?
The authors find that CDS coverage significantly increases the probability of loan delinquency and that loans originated after CDS coverage was put in place have a much higher likelihood of becoming delinquent than loans originated before CDS coverage was initiated. Their analysis also shows that the largest CDS effect is for loans securitized by commercial banks. The authors speculate that these banks kept the highest-quality loans on their books or allocated these loans to MBS deals that were not insured while allocating the riskier loans relying on soft information on borrower credit quality to the MBS deals that were insured by CDSs. They found a strong CDS timing effect across all of the major loan securitizers (i.e., commercial banks, investment banks, and finance companies that specialized in loan origination).
Furthermore, although CDS coverage is found to increase the probability of loan delinquency when the collateralized debt obligation settlement date is no later than 180 days after the MBS closing date, the effect is more profound when the authors test a narrower window—for example, when the settlement date is within 90 days before the MBS closing date. They discover that the state-level change in the unemployment rate has a negative effect on loan delinquency instead of the predicted positive effect but suggest that lenders in states with higher unemployment rates might have been more careful when issuing loans.
How Did the Authors Conduct This Research?
Multivariate probit models are used to test various predictions and sample periods. A sample of 9,606,797 privately securitized subprime mortgages originated during 2003–2007 is derived from a database constructed by First American CoreLogic LoanPerformance, which contains more than 90% of the subprime loans privately securitized during the sample period. The database provides such information as the loan origination date, information on the mortgage loan pool, the securitizer, the MBSs holding the loan, information on the borrower, and other loan characteristics. The authors collect data from other sources related to regional housing and the economic conditions at the time of loan origination. They find that more than 35% of the subprime loans in the sample were in pools covered by CDS contracts concurrent to the closing date of the MBSs that contained the pool.
The authors control the results for various scenarios that could explain the results, including that the findings are the result of using CDS contracts to hedge the risk of already outstanding loans and that the findings are explained by CDS coverage in regions or time periods with high mortgage defaults. In both cases, loan delinquency is significantly affected by CDS coverage, with the exception of the loan group with the lowest percentage of CDS coverage. Loan delinquency is used as the proxy for subprime loan performance, and a loan is defined as delinquent when it is at least 60 days past due within the first 24 months of origination.
This research takes a fresh look at the market dynamics that led to the financial crisis of 2007–2008 and challenges a widely held belief that looser lending standards associated with the originate-to-distribute model were the primary reason for the massive residential mortgage defaults. By examining the relationship between CDSs and MBSs, we are closer to a better understanding of the behaviors and events that led to the crisis. Hopefully, this understanding will help prevent a repeat of history.