How Scores Shift in Recessions

Commercial Scores

The Great Recession vs. The COVID-19 Pandemic

PayNet, an Equifax Company SVP & General Manager Bill Phelan and Senior Data Scientist Amanda Hull provide background on how scores shifted during the Great Recession and compare those shifts to how scores have been impacted by the 2020 COVID-19 pandemic.

The average MasterScore® v2 (MSv2) score between the Great Recession and the recovery differed by as much as 20 points. Typically during a recession, portfolio average scores initially trend downward due to increasing delinquency, and then trend upward as lenders tighten their belts and lend to only higher-credit quality businesses, and at-risk businesses go out of business and drop from the borrower pool. A survival-of-the-fittest impact can be seen, where average scores at the end of a recession exceed those prior to the recession. Those businesses still active at the end of the recession are those with the highest resilience, and therefore higher scores.

Overall default rates in the depths of the Great Recession were 2-3.5 times higher than default rates in the recovery. But by MSv2 band, default rates were typically only 1.5-2 times higher; the difference is less significant by score band due to MSv2 absorbing some of the economic shock and adjusting the score values accordingly – but there is a default rate shift nonetheless.

The highest-scoring borrowers have very little default rate shift – scores 720+ were reliably “good” even in the Great Recession. But lower scores had more significant default rate shift. Lower scoring businesses are more at-risk during a recession and show greater odds of shifts in recessions. Higher scoring businesses are more resilient and show little shift in odds during recessions, particularly in models that incorporate macroeconomic information.

This new economic crisis is unprecedented; while it won’t necessarily follow score/default rate patterns that we saw in the Great Recession, we can expect it to be as bad or worse, at least for the hardest-hit segments of the economy, like retail businesses, healthcare, and transportation. Lenders using score thresholds in their processes should adjust those thresholds, especially if the thresholds are set up to align to an expected default rate, as default rates at a given score value can be expected to increase as they did in the Great Recession (or more). Lenders with more complex scoring models should consider a model recalibration tuned towards a recessionary scenario, or at minimum an adjustment of how the output of the existing model is used. This is also a great time for lenders who have not done so to implement regular monitoring of score/model trends so that they can stay on top of shifts in real time.

PayNet can provide additional data to help lenders adjust, whether industry-wide trends or a lender-specific dataset. Lenders should also try to stay on top of economic trends to align their processes with the marketplaces. They can do so by monitoring the PayNet Small Business Indices, which serve as a leading indicator of macroeconomic and industry trends.

In conclusion, as lenders are making adjustments to their use of credit scores for COVID-19, here are key considerations:
- Scores will likely start decreasing if they haven’t already; scores that react to economic factors will decrease more.
- At the same time, the expected default rate for a given score value should be estimated as at least 1.5-2 times higher than the previously expected number.
- Portfolio score trends should be monitored, as should industry/economic trends, at least monthly.

Download the PDF

For more information on our commercial credit scoring data, contact us.