Assessing borrower risk during COVID-19 — a six step action plan

Eight months into the COVID-19 pandemic and there is still a high degree of economic uncertainty and anxiety. Jerome Powell, Federal Reserve Chair, recently told a congressional panel: “America’s economy has shown marked improvement since the coronavirus pandemic drove it into recession, but the path ahead remains highly uncertain.”

According to an updated CUNA forecast, there are additional crisis threats outside of the pandemic, like west coast forest fires, east coast hurricanes, US political tensions, and disappearing government benefits that could impact credit union earnings. One thing is sure, COVID-19 will not be the only crisis lending institutions face over the next year.

There are a few bright spots: share of mortgages in forbearance dropped from 7.2 percent to 6.93 percent as of September 13, according to data from the Mortgage Bankers Association. However, the trade group also revealed that not all segments of the market have improved.

“While housing market data continue to show a quite strong recovery, the job market recovery appears to have slowed, and we are seeing the impact of this slowdown on FHA and VA borrowers in the Ginnie Mae portfolio,” said Mike Fratantoni, chief economist for the Mortgage Bankers Association.

‍

COVID-19's impact on credit models

Amidst the ongoing pandemic and increased likelihood of continued uncertainty due to other crises, lending institutions are still figuring out how to evaluate and monitor credit risk with limited visibility and access to reliable data.

Cornerstone Advisor’s 2020 state of credit modeling report revealed the following:

 
• Only 10 percent of financial institutions with more than one billion dollars in assets (and 20 percent of smaller lenders) say they’re “very confident” in their models’ abilities to maintain their effectiveness post-COVID
   
• Only 15 percent of lenders with more than one billion dollars in assets said they are “very prepared” to adapt
   
• Most lenders, especially the bigger ones, admit that it’s somewhat or very difficult to make updates to their credit models, a process that can take many months
  ‍

Assessing borrower risk with a six step action plan

In the current environment, lenders can’t rely on credit scores as they previously have. So, what approach would enable lenders to get more predictive about a consumer’s ability to repay, beyond their credit score? We’re glad you asked. Here at Zest AI, we get to work with leading financial institutions and credit unions and recently presented how a customer used the following approach with their AI-powered model to increase visibility and make better decisions faster:

Step 1: Understand how scores are changing — are they lower? Higher?

Step 2: Uncover and scrutinize changes in borrower characteristics.

Step 3: Understand how attributes are influencing model scores.

Step 4: Understand your anomaly rate — how different are the current applicants?

Step 5: Use alternative data and apply more scrutiny when warranted/

Step 6: Rigorous portfolio review to quantify repayment risk.