Credit unions and community banks face different CECL challenges
Community banks and credit unions face different challenges when preparing for the new current expected credit loss model, also known as CECL.
Often, credit unions are grouped together with community banks, although their experience will look different when building out their models. What loss rate methodologies are applicable to community banks, but not credit unions? What challenges do credit unions face that are unique to their own loss history, portfolio structure and loan count?
It’s likely that data challenges are so great for community banks that certain loss rate methodologies cannot be utilized. With credit unions, this is not the case. Most loss rate methodologies make sense for a credit union’s portfolio. A more common challenge for credit unions is a lack of careful steps when putting the methodologies together. When making the calculations, staff should be considering the right elements and watching out for things that can be punitive to the overall allowance and introduce volatility period over period.
“While most of these CECL methodologies may be perfectly fine for credit unions, we first must have the data to support it,” said Jared Mills, a senior consultant at Sageworks. “That’s the challenging thing: We never had to have this data historically; we didn’t have to have this loan-level detail for the allowance. As of 2016, things changed. Now we need a lot of this data to support these methodologies.”
CECL expects financial institutions to segment loans into pools with similar risk characteristics that are as granular as possible while maintaining statistical significance. Community banks often face a challenge of having a low number of loans in each of their pools, to the point where achieving statistical significance is difficult. They can have large, commercial portfolios with significant balances but insignificant loan counts. This issue is often a topic of discussion that community banks seek feedback on. However, this is typically not a concern for credit unions. In an informal poll of over 63 credit union managers and executives during a recent Sageworks webinar, none stated that having a low number of loans in each pool is a challenge relevant to their institution because of a lack of commercial real estate lending.
Having sporadic, unpredictable patterns of loss is a more common trend that credit unions observe. Mills explained that it is a case in certain portfolios, especially recently, where there hasn’t been a charge-off in a while. One event could potentially throw the pattern off in a large manner.
In a recent “Ask the Regulators” webcast, regulators from the Federal Reserve System, the Federal Deposit Insurance Corporation, the Securities and Exchange Commission, the Conference of State Bank Supervisors and the Financial Accounting Standards Board discussed common loss rate challenges for community institutions. However, they did not differentiate between which challenges might affect community banks and which might impact credit unions. Mills provided comments on these points. Below, in bold, are his comments on whether each challenge is something that typically pertains to credit unions:
• Losses are minimal – Potentially
• Losses are sporadic with no predictive patterns – Potentially
• There is a low number of loans in each pool – Not typically, but having enough granularity is more of an applicable topic
• Data is only available for a short historical period – Often
• Today’s portfolio composition varies significantly from historical portfolios – Potentially
• There are changes in the economic environment – Often
“For credit unions, the gold standard of the allowance, in general, is to keep our models stable under stable conditions. Keep it responsive under changing conditions,” Mills said. “The point of CECL was that current models do not adapt in the way that they need to, so we need to evaluate ahead of time that our models do change. This includes hypothetical conditions when we are not in that particular circumstance.”