The banking industry is about to find out what is “reasonable and supportable” when it comes to predicting future economic conditions and the loan losses that follow.
The new credit loss standard requires banks to create and incorporate a “reasonable and supportable” future forecast of economic conditions into their loss estimates. The broad and vague requirement has become another hurdle for bankers as they implement and comply with the new approach to loan loss accounting. Disclosures from banks with 2020 effective dates show a diversity in what management teams believe is “reasonable and supportable,” which could make it challenging for community banks wondering what to do.
A future economic forecast is critical to the current expected credit loss standard, or CECL. Banks must forecast the lifetime losses for their loans, ergo they must have some sense for how future conditions could change the performance of those loans. The forecast should be “reasonable and supportable;” after the forecast period, banks can revert to their historical loss data.
Already, there is a variety in how far banks will forecast into the future, according to a sampling of large banks’ third-quarter filings. KeyCorp, which has $146.7 billion in assets and is based in Cleveland, expects to use a two-year forecast period. Green Bay, Wisconsin-based Associated Banc-Corp, which has $32.6 billion in assets, and Evansville, Indiana-basedOld National Bancorp, which $20.4 billion in assets, will both use a one-year forecast period. U.S. Bancorp will use a three-year forecast period and multiple scenarios. PNC Financial Services Group will incorporate “four macroeconomic scenarios and their estimated probabilities” and use a forecast period of three years.
“These will be produced by our economics team using a combination of structural models and expert judgment and will be designed to reflect a range of plausible economic conditions and emerging business cycles over the next three years,” PNC says in the filing.
But what is “reasonable and supportable” when a community bank is trying to predict what will happen in its markets to its loans? The accounting board behind the standard released an Q&A over the summer to address concerns about creating the forecast. Auditors and accountants say the guidance is helpful, but point out limitations for banks struggling with execution.
The Q&A shows that the board does not expect or require community banks that use “a fairly simplistic model to, now, do something really complicated … in order to comply,” says Will Neeriemer, a partner in the Financial Institutions Services Group at DHG. Bankers may be able to use it to defend their judgement and explain the appropriateness of simplistic models to regulators and auditors.
What the Q&A doesn’t do, however, is lay out a roadmap for forecast creation or outline appropriate methodologies. Like many things about CECL, creating a reasonable and supportable forecast is principles-based and requires each bank to come up with their own approach, with their own inputs and data.
“I think that’s where community banks are struggling: They’re stuck with having to come up with a less-complex model on their own,” Neeriemer says.
One way for community bankers to begin creating a forecast is reframing it with the question: “‘Why did I reach the judgments that I reached? Can I explain this to a third party?’” says Graham Dyer, a partner in the Accounting Principles Group at Grant Thornton.
For instance, he says executives can adjust their historical loss information to look for differences in performance and loss rates during various economic cycles, comparing them to current conditions and forward-looking economic data. He points out that banks make credit loss estimates today under the existing allowance approach, and discuss and defend those judgements with external auditors and accountants. “There’s always a reason,” he says. “It doesn’t always have to be a highly sophisticated mathematical model. But it probably does require some data.”
Takeaways and Insights for Creating a Reasonable and Supportable Forecasts:
Neeriemer says that community banks that use a qualitative factor framework — sometimes called Q-factors — that has been audited and accepted by regulators and their auditors could leverage or incorporate that into their forecasts, if appropriate.
Neeriemer has seen community banks incorporate broad, public data into their forecasts, like the unemployment rate, along with financial metrics like interest rates and the yield curve. Both Neeriemer and Dyer emphasize the importance of a bank’s own historical loss data. This is often a bank’s most accurate and useful information when trying to figure out how loans performed in different economic cycles, and may indicate if an institution’s performance runs counter to the direction of the economy. For instance, a bank may learn that the unemployment rate may not be a useful metric for forecasting if its losses don’t seem to move in the same direction as the rate.
Which Way, How Much
Dyer says there are two factors bankers should think about when making adjustments to their lifetime allowance: directional consistency and quantitative sufficiency. When economic conditions change, does the change in their allowance mirror that, or go against it? How much does it change?
Dyer recommends bankers have early and frequent communication with regulators and external advisors as they create and update their forecasts, so outsiders have adequate opportunities to respond.