Everyone's been hit in some way by covid-19. Lenders have been hit as well. The graphs below show the loan loss reserves for 3 lenders as a percent of credit card loan balances at the end of 2019 before CECL adoption, the impact of CECL effective on January 1, 2020, and the increases in loan loss reserves in each of the first 2 quarters of 2020. Needless to say, it's not pretty.

Capital One and Bank of America information is for their domestic credit card businesses. For American Express it only includes cardmember loans.

Yes and No. CECL's intent was to quantify the risk in the portfolio at a given date. Implicit in that is a set of assumptions on the health of the economy and how it will impact customer delinquency and losses. Clearly the 2020 covid-19 recession saw a dramatic spike in unemployment. Small business are hurting with many likely to go out of business. It's hard to know whether banks have been able to capture all of the risk in their portfolio given that we don't know how long the recession will continue until things return to 'normal'.

It's unlikely anyone was forecasting a recession like 2020's. But it still seems that economic assumptions should stress test a range of outcomes. While not likely it was always possible. Just not likely. Or not at all likely. But with some bank's reserve percentage almost doubling it seems they didn't take into account much of a downturn occurring at any time over the life of their portfolios. That seems to have missed the point.

The idea of quantifying the risk in the portfolio still makes sense. But it does come down to what assumptions are reasonably expected. It's hard to judge the initial CECL reserve setting given we have the benefit of retrospect. But it's a lesson that other lenders should take into account as they adopt CECL over the next several years.

Here are links to the 10-Q reports used above: