The American Institute of Certified Public Accountants (AICPA) held its National Conference on Banks & Savings Institutions virtually, on Sept. 14-16, 2020. SS&C Primatics was once again honored to be a Silver Sponsor and to have participated in the speaker series as subject matter experts on CECL. Respected industry thought leaders discussed the key areas of focus at this year’s conference, the current expected credit loss (CECL) model under ASC 326, Financial Instruments - Credit Losses, and the impact of the pandemic.
Among the recognized industry thought leaders was SS&C’s SVP of Product and Operations, John Lankenau, who kicked off day two of the conference session with the session “Navigating CECL in the COVID-19 World.” Lankenau spoke to a full house on how the industry’s view of the allowance under CECL has changed over time. His topic subtitled, “What Have we Learned and What will Future Challenges Be?” dove into the transition to CECL during the pandemic.
Attendees heard how financial institutions have transitioned to CECL and current thought leadership for the journey ahead. Lankenau outlined the impact of the CECL adoption during these unprecedented macroeconomic times. Those institutions that saw CECL as a purely modeling exercise struggled; as Lankenau reminded the audience, CECL is an accounting standard that requires quantifying credit losses over the contractual life of the instrument. And this has presented many data challenges because:
- CECL relies on different data elements than old reserving processes. For example, contractual terms
- CECL requires the life of loan to be described accurately. For example, construction to permanent
- Legal interpretations matter. For example: conditionally vs. unconditionally cancellable
- Rate/Volume analysis is no longer conceptually correct, so more data is needed closer to quarter end, as CECL is a more complex process that requires higher data quality across more elements
Often, the lack of availability of this type of data has led institutions towards choosing a specific methodology (i.e., open pool), which may be less granular than initially desired.
Lankenau then defined the oft-confusing term “model” and focused on what makes a model useful and fit for purpose. While noting many factors can compromise the integrity and usefulness of a model, he presented some requirements for a good model:
- Transparent and institutive: quantifies the relationships that you already know
- Appropriate techniques: uses well-understood methodologies that are fit for the purpose
- Manageable and useful: has enough variables to quantify a relationship, but not so many that it’s not maintainable or understandable
- Accurate and reasonable: on average, correct with reasonable answers over a range of plausible inputs
- Implemented accurately, used appropriately and understood by users
Finally, because of COVID-19 and the policy response—both fiscal and monetary—we are in unprecedented economic times, and that will test models and potentially hurt their ability to be useful. While model validation and governance have always been important, the focus on this part of the process will become even more important as models are being stretched and the usefulness of these models is coming into question.
This has been a challenging time for banks, as they are facing implementing new credit loss standards amid unprecedented economic upheaval. In order to ensure that credit loss estimates are appropriate, banks should ensure that all required data is available and validated, that models are fit for purpose and that results are intuitive. As this will take time and specialized expertise, banks should plan accordingly.
Access the full video of John’s session on Navigating CECL in the COVID-19 World: What Have we Learned and What will Future Challenges Be? here.
Explore SS&C’s library of CECL resources to learn more about modeling for CECL.