Experts Comment on Banker Responses to CECL Poll
Almost half of U.S. financial institutions have yet to begin preparing for CECL. But many more than half agree they don’t have the data they’ll need to be prepared and that they’ll need something more capable than their current Excel spreadsheets to estimate their allowance under CECL.
Those were some of the results of poll questions posed to more than 300 bankers who registered for the March 21 MST webinar on a seven-step approach to developing and implementing a CECL-compliant allowance model. During the webinar presented by MST Executive Vice President John Closs, webinar attendees were asked to respond to five poll questions relative to their CECL preparation and expectations. As a follow up, MST asked five industry experts to comment on the poll answers. These industry thought leaders are among those presenting at the National ALLL Conference held May 24-26, 2016 in La Jolla, California.
Poll Question #1: Have you begun preparing for CECL?
Banker Responses:
Yes – 58%
No – 42%
Expert: 
Molly Curl | Partner, Bank Advisory and Regulatory Services | Grant Thornton, LLP
“I’m glad to see that the majority have at least started to prepare. I would have liked to see this as a higher percentage as the implementation will be upon us before you know it. The reason we should start preparing is that the more concrete, supportive historical information that can be obtained the better. Consider for example the information you would need to support the loss rate used for the life of the loan on various loan product types. If you have started collecting the losses incurred for the life of a loan currently you will have fewer assumptions that you will need to support with outside statistics that may not be as favorable as your own historical performance.”
Poll Question #2: Do you have access to all of the data that you’ll need to comply with CECL?
Banker Responses:
Yes – 39%
No – 61%
Expert: Graham Dyer | Senior Manager, National Professional Standards Group | Grant Thornton, LLP
“The results are in line with my expectations, that most banks believe they do not have the data they will need to implement CECL. However, we do think there are approaches that all banks will be able to utilize initially as they build the datasets they will need on a long-term basis. The message this result should send loud and clear is that if your institution is among those that answered “no,” then you need to start the data collection process immediately.”
Poll Question #3: What reserve methodology are you likely to implement for CECL?
Banker Responses:
Historical Loss – 63%
Migration analysis – 36%
Cash flow analysis – 1%
Expert: Chris Emery | Director of Special Projects | MST
“This is consistent with what we’ve heard from FASB, as well as what we’ve heard from our clients and other institutions with whom we have discussed this issue. Although FASB’s initial expected loss forays seemed to specify a DCF (discounted cash flow) approach, the December 2012 exposure draft made it quite clear that would not be a requirement. In addition, based on what FASB has said publicly, they fully expect many banks to be able to use their current models as a starting point for eventual CECL implementation. For a vast majority of banks, their current model would not be based on a DCF approach for most loans.
“It is important to point out that although a current model may be a starting point for CECL implementation, it will almost certainly not be the ending point. Annualized loss rates that are the bases for many bank’s current ALLL models will not be sufficient, nor will it be sufficient to simply multiply those annualized loss rates by the average life of the pool of loans it is being applied to (see paragraph 825-15-55-24 in the exposure draft). So while banks may anticipate using a model based largely on historical losses, they will almost certainly need to adjust their method of arriving at as well as how they apply those established historical loss rates.”
Poll Question #4: Do you think CECL will . . .
Banker Responses:
Increase reserves – 74%
Decrease reserves – 5%
No change in reserves – 21%
Experts: Mike Thronson and Gabe Nachand | Partners | Moss Adams, LLP
Mike Thronson
“It depends on how much a bank has in its current allowance. For instance if they have large unallocated and QE reserves it may not increase the ALLL very much. If they have small QE and unallocated reserves an increase is more likely. That said, the bank should not bolster its reserves between now and implementation.”
Gabe Nachand
“Based on what we are seeing and hearing related to those modeling CECL, reserves are definitely increasing, mostly because of the increased loss horizon as a result of an expected credit loss model (expected life) versus a shorter loss discovery period being utilized or embedded within current methodologies under the inherent loss model. A couple opportunities for institutions to mitigate this impact relate to loan lives and portfolio segmentation. Being able to support a portfolio loan life of less than contractual maturity will likely reduce the overall reserve estimate for many of the methodologies under development. Segmentation can also significantly impact the estimate. Inadequate portfolio segmentation can result in historical losses in very narrow loan categories being applied to larger loan pools that really aren’t similar in credit risk, resulting in inflated reserve estimates.”
Poll Question #5: Do you plan to use Excel or a formal vendor solution to manage CECL?
Banker Responses:
Excel – 36%
ALLL software solution – 64%
Expert: Michael Gullette | Vice President, Accounting and Financial Management | American Bankers Association
“With a third of respondents saying they will stick to spreadsheets under CECL, I think it will be interesting to see what methods bankers plan to use to estimate loan losses, both in the short term and long term.”
Registration closes April 30, 2016 for the National ALLL Conference.
Download the Seven-Steps to CECL-Compliance whitepaper.
