Default Option Exercise over the Financial Crisis and beyond
Xudong An, Yongheng Deng, Stuart A Gabriel
Review of Finance, Volume 25, Issue 1, February 2021, Pages 153 – 187, https://doi.org/10.1093/rof/rfaa022
Default on residential mortgages skyrocketed during the late-2000s, giving rise to widespread financial institution failure and global financial crisis. Among factors associated with mortgage failure, analysts have pointed to factors including widespread incidence of negative equity, shocks to unemployment and income, lax underwriting, expansive use of risky loan products, and mortgage fraud. In this paper, we provide new evidence of heightened borrower sensitivity to negative equity and show that factor to be highly salient to crisis period defaults. The elevated borrower default response to negative equity, coupled with a decline in home equity, resulted in a widespread increase in mortgage default during the 2000s crisis. Results also suggest elevated default option exercise by borrowers seeking to qualify for crisis-period loan modification programs.
To identify changes in borrower response to negative equity, we estimate a time-varying coefficient competing risk hazard model to loan-level event-history data. We model the conditional probability of default as a function of contemporaneous borrower negative equity and a large number of other factors. We label the estimated coefficient associated with negative equity the “default option beta”. Results of rolling window local estimation of the hazard model show a marked run-up in the default option beta from 0.2 during 2003-2006 to about 1.5 during 2012-2013, followed by a downward retracing of roughly one-half the upward movement in the estimated beta value through 2016 (Figure 1). The upward movement in the default option beta led to substantially higher default probabilities for a given level of negative equity. Model simulation shows that the marked upturn in the default option beta resulted in a doubling of default incidence. Results further indicate substantial geographic heterogeneity in the default option beta.
We also find that county unemployment rate shocks, reflecting cyclical fluctuations in the local economy, are highly predictive of variation in the default option beta. Conditional on controls for the local business cycle, we find that borrower default propensities are sensitive to consumer distress. Further, those factors are economically salient and together could account – via their impact on the default option beta – for over two-thirds of the increase in crisis period default risk. We further find evidence of a structural break in the default option beta time-series in 2009, which coincides with federal mortgage market intervention via the Home Affordable Modification Program (HAMP). Also, while results do not show significantly damped default propensities among states with recourse to borrower non-housing assets, they do indicate sizable and significantly elevated default option betas among sand states hard hit by the 2000s housing and mortgage crisis.
Finally, while beyond the timeframe of our study, findings suggest substantial evolution in borrower response to negative equity in the wake of the 2020 COVID-19 pandemic and related mortgage policy interventions. We leave that analysis to future research.