A new accounting standard will affect credit unions much more severely than community banks, according to a study to be released Oct. 17.
The research, a joint effort between Prescient Models and Deep Future Analytics, also shows community banks could release 59% of their reserves in adopting the current expected credit loss, or CECL, standard. By contrast, credit unions would need to increase their reserves by 22%.
A massive release runs counter to several forecasts that banks would need to build reserves to comply with CECL. However, the study's model does not assume a recession, and most economists think one will hit within the next two years. The model uses the baseline scenario from this year's regulatory stress testing exercise for the two-year "reasonable and supportable" forecast before reverting to a long-term historical mean, as recommended by the accounting board behind CECL.
Re-running the model using the stress test's severely adverse scenario for the "reasonable and supportable" two-year forecast shows much different results: A reserve build of 69% for community banks and 107% for credit unions. The regulators' severely adverse scenario incorporates a steep recession with gross domestic product growth declining by 9% at one point.
Even if a bank's macroeconomic forecast and CECL model suggest a reserve release is appropriate, banks are unlikely to follow through with a significant release, said Joseph Breeden, founder of Prescient Models and an author of the study. He said most banks would cite qualitative factors to simply maintain reserves at current levels.
"Was the point of CECL to have a model say your reserves should be lower and have everyone override it?" Breeden asked in an interview.
The study used data from 103 clients, roughly split between banks and credit unions. The study is a parallel run that presumes the depositories adopted the standard in August. Most small financial institutions do not have to implement CECL until 2023.
Breeden said the divergence between credit unions and community banks is likely a reflection of the differing aims of their respective regulatory agencies. Whereas the Federal Deposit Insurance Corp. conservatively presses banks on safety-and-soundness to avoid hits to the agency's insurance fund, the National Credit Union Administration can view significant reserves as a negative, since credit unions are nonprofits that should deploy excess capital.
Breeden said the study confirmed the central role of product life in CECL, as loans with longer terms, such as mortgages, required significantly higher reserve builds. The study also underscores the role of macroeconomic factors and suggests depositories might want to use multiple scenarios given the checkered history of economic forecasting.
"You're not required to run more than one scenario, but if I were a CFO, I would want to see both answers. I think the severe is too severe and the base is too optimistic," Breeden said. Companies could rely on consensus economic forecasts, Breeden said, while cautioning that economists are often too optimistic.
"There's been recent studies that show actual recession is usually a year before they think it was, so if the consensus says a recession will be in two years, I'd go with one year," he said.