Unfortunately, the “triumvirate of hurricanes” – Houston, Florida and Puerto Rico have given banks a modern data set in which to model a widespread disaster scenario for their residential first trust deed credit portfolio. While every disaster is different, these three events were similar enough to be comparable but different enough to create some modeling variability to make it useful. Further, while Puerto Rico is a slight outlier, Houston and Florida almost perfectly line up with an average American credit and so presents a statistically useful benchmark to see a before and after snapshot. If your bank’s portfolio is susceptible to hurricane, floods, earthquakes or other disaster impacting a county-wide area or more, this data will provide some support to help you model what an “event shock” could look like.
While these disasters caused many loans to go delinquent due to payment interruption and dislocation, more than half have brought themselves current. 5.4% of the loans prepaid above normal levels. Applying a 30% loss given default, or average severity metric, cumulative losses are approximately 53 basis points (bps) of the impacted loans. If banks had to liquidate the remaining loans that are still delinquent, there would be between two bps and five bps of additional losses in the portfolio.
Putting This Into Action
The good news is that these performance rates are not catastrophic to banks by any means despite the catastrophic impact of the hurricanes. In addition to credit shocks on the portfolio, prudent risk managers will want to create an event shock customized for the disaster scenario applicable to their service area. This experience gives banks a great set of benchmarks to do just that.
Submitted by Chris Nichols on May 07, 2018