The Calm Before the Credit Risk Storm: Following Up on Lessons From Other Natural Disasters

Four months ago, AQN examined whether lessons from natural disasters like Hurricane Katrina could help banks forecast longer-term economic impacts on their portfolios.  At that time, the unemployment rates of the US at the start of the COVID lockdown mirrored the early unemployment rates for Louisiana post-Katrina.  We asked the question then, “Will the US during COVID-19 follow the path of Louisiana longer term?”.  So far, that has not happened.  One year after Katrina, the average New Orleans consumer saw their debt decreased by $8.5K, 90+ DQ rates increased by 10 percent, and risk scores dropped by 4-7 points.  But going into September of 2020, the story is very different.  While US unemployment remains at record highs, credit risk seems at an all-time low.  So, what is going on?

For lenders, the credit risk storm clouds are likely just appearing on the horizon.  Most banks saw decreases in origination volumes and significantly increased their loss reserves back in Q1 and Q2. The CARES act appears to have softened the initial blow, but enhanced benefits have ended, and it’s unlikely new legislation passes before November.  For the next six months, AQN expects to see increases in risk on consumer debt and a differentiated response from lenders who successfully, or unsuccessfully, navigate the COVID-19 bubble.  The first key differentiator for lenders will be whether they see it coming.

When layoffs began in March, most banks offered 3-month payment deferral programs for consumer and SMB loans.  This obfuscated the impacts of a consumer’s ability to pay as banks were not actively attempting to collect on these hardship debts.  In fact, between the first and second quarter of 2020, early stage delinquencies decreased 20% from an already 15-year low.

A 20% decrease in Q2 0.00% 0.50% 1.00% 1.50% 2.00% 2005 2010 2015 2020 Quarterly Report Date DQ Rate Delinquency 30-89 DPD 90+ DPD DQ Rates - Quarterly Call Reporting

Source: Call Reports data provided via bankregdata.com

 

The decrease, driven by hardship programs and enhanced unemployment benefits, is at odds with unemployment data which shows a much bleaker picture. Relative to Katrina the number of claims is roughly double, peak unemployment lasted nearly twice as long, and the tail seems to be on-going at a rate of nearly 50K more claims per week than at peak Katrina.  At the same time, many bank risk models are seeing their portfolios at the lowest risk they have ever been.  The combination of lower card utilization and lower credit inquiries are making customers look lower risk.  The question lenders should be asking themselves is, “Do I really believe the CARES act turned my risk bubble into a windfall?”  If you answered “Yes” remember that 90+ DQ rates increased 10% after Katrina; they did not remain at historic lows.

Week 5 - Stimulus Payments Start Week 18 - Enhanced Unemployment End Week 2 - CARES Act Week 0 - National Emergencies Declared 0 100,000 200,000 300,000 400,000 -5 0 5 10 15 20 Weeks Since Event Unemployment Claims Event Hurricane Katrina COVID-19 Louisiana Weekly Unemployment Claims - Katrina vs COVID-19

Banks are likely on the precipice of seeing a two-fold impact in their current credit risk reporting.  First, consumers have started aging out of the first round of hardship programs and second, CARES benefits have expired, including enhanced unemployment.  From AQN’s work with both FinTechs and banks, we have seen multiple firms struggling to monitor these time-series impacts because of the same core problem:

  • Technical teams were asked to rapidly develop new solutions to implement hardship programs

  • Hardship variables were stored in locations not accessible to analysts responsible for monitoring

  • Data was not in formats easily consumable by existing monitoring dashboards and reports

  • As customers graduate or roll into new hardship programs, data is often conflicted reducing trust in the data

These problems have led to time-delays and blind spots in understanding the COVID-19 bubble.  Those still struggling to accurately monitor hardship and delinquencies could have a diminished ability to collect on debts now that the bill is due and unemployed consumers have $600 a week less entering their bank account.  In hindsight, the great wave of payment deferrals may have been mistimed.  They might have been better implemented now as enhanced unemployment benefits are expiring rather than in the spring of 2020.  Most finance teams have already prepared for the fallout by adjusting their loss forecast but there is wide variability across banks - a likely indicator that some feel their ability to monitor and react will be a key differentiator relative to their competitors.

To help your lending business prepare for the upcoming bubble, please reach out to AQN for examples of how we are helping our clients navigate the impacts of COVID-19.  And be sure to also check out how AQN is recommending lenders think about collections and no-regret pandemic responses. 

Guest User