Skip to main content

Looking for Valuant? You are in the right place!

Valuant is now Abrigo, giving you a single source to Manage Risk and Drive Growth

Make yourself at home – we hope you enjoy your new web experience.

Looking for DiCOM? You are in the right place!

DiCOM Software is now part of Abrigo, giving you a single source to Manage Risk and Drive Growth. Make yourself at home – we hope you enjoy your new web experience.

How to Forecast Future Expected Credit Losses

Mary Ellen Biery
October 2, 2015
Read Time: 0 min

For an updated and more detailed explanation of forecasting future expected credit losses, see the post, “Forecasting: Considerations for a Reasonable and Supportable CECL Forecast


A central difference between the existing incurred-loss model for estimating credit losses and the FASB’s proposal utilizing a current expected credit loss, or CECL, model is that financial institutions will need to estimate credit losses on loans over the life of the loan.

Shifting from estimating only losses already incurred to forecasting future expected credit losses raises the question of how institutions can generate what standard-setters call “reasonable and supportable” forecasts. This question is a major one for U.S. banks and credit unions, according to Graham Dyer, senior manager in Grant Thornton’s National Professional Standards Group.

Under CECL, Dyer said recently at the 2015 Risk Management Summit presented by Abrigo (formerly Sageworks), “We’re required to make forecasts of the future – that’s the big hang-up.”

Dyer, who also serves on a task force related to implementing similar accounting rules internationally, said regulators in the U.S. and abroad see a strong tie between sound credit practices and the estimation of credit losses. Because credit risk personnel are most likely to have data that links events with credit losses, involving credit risk management in the CECL transition will be critical, he said.

“There’s a large degree of estimation uncertainty here and everyone understands that,” Dyer said. “I don’t think anyone’s saying you have to, with pinpoint accuracy, determine how events will definitely impact credit, but you need to be able to draw a reasonable and supportable line between the conditions being forecasted and credit losses.”

Learn more about navigating the CECL transition.

Request More Information »

He also reassured bankers that the process may not be as complicated as they fear. “I’m not trying to say there’s not going to be effort involved; there will be,” he said. “For larger, more complex institutions or even institutions with complex credit exposure, this is going to be difficult. But for a lot of places, this may be not as hard as it seems.”

Here is a sample six-step process outlined by Dyer for forecasting future expected credit losses:

  1. Identify pools of credit exposure with shared risk.
  2. Identify, by pool, the primary drivers of credit risk. “What are the economic conditions we think are pretty reasonably correlated to credit losses?” Dyer said.
  3. Correlate changes in the primary drivers to credit losses. Dyer said this part of the process doesn’t require a PhD in mathematics. “There’s some math,” he said.  However, “A lot of this can be done on a qualitative basis with corroborating evidence. Our credit risk folks, our loan workout guys, they tend to know what drives that risk. They’ve got experience doing this; they know when losses are coming, so go talk to them. Say, ‘I’ve got this commercial real estate portfolio. What do you guys look at when you think about future credit risk for this portfolio?’” It could be a metric like localized GDP or something from the Fed’s Beige Book. Dyer suggested looking at two or three metrics and obtaining evidence that corroborates how those are correlated to losses. A fairly simple analysis, in many cases, can show that the losses and the metrics tend to move in the same direction, and according to a consistent correlation, Dyer said. “We can perform linear regression, or even more complex analysis, if necessary, but you might be able to do something as easy as that in a relatively simple environment.”
  4. Forecast a future expectation of the primary drivers of credit risk over a reasonable and supportable period. Dyer said this step shows the importance of involving early on the right people to ensure the appropriate data is available when you need it.
  5. Use those correlations to estimate future expected credit losses for a reasonable and supportable time period. For example, an institution might determine a reasonable estimate for GDP would go out a year, so it would apply that forecast for one year of the remaining contractual period.
  6. Beyond that reasonable and supportable period, institutions could presume that the correlating metric would revert to the long-term average, or mean.

Following a process similar to this will allow credit risk managers and finance staff to work together to begin forecasting future expected credit losses. For more practical advice on transitioning to an expected-loss model for measuring credit impairments, visit

About the Author

Mary Ellen Biery

Senior Strategist & Content Manager
Mary Ellen Biery is Senior Strategist & Content Manager at Abrigo, where she works with advisors and other experts to develop whitepapers, original research, and other resources that help financial institutions drive growth and manage risk. A former equities reporter for Dow Jones Newswires whose work has been published in

Full Bio

About Abrigo

Abrigo enables U.S. financial institutions to support their communities through technology that fights financial crime, grows loans and deposits, and optimizes risk. Abrigo's platform centralizes the institution's data, creates a digital user experience, ensures compliance, and delivers efficiency for scale and profitable growth.

Make Big Things Happen.