It takes at least two people to waste time reinventing wheels; that kind of inefficiency requires multiple departments to work outside of their day jobs – certainly an increased risk in these distanced times. Even the most communicative organizations can still spring a wasted effort trap if similar problems are not recognized as sufficiently “wheel-like.”
Activities from pricing, budgeting, allowance preparation, and stress testing depend on current and forward-looking expectations for the volume and timing of credit losses. Estimating losses in an environment that differs from recent historical experience in a consistent, quantitatively justified manner requires the use of some form of modeling approach. While the application of those models and underlying assumptions vary by activity, the model should still reflect the institution’s best estimate, and should not consider any reliable inputs “off-limits” during development.
Thoughtfully developed models can substantiate qualitative adjustments for institutions still using the incurred loss method for the allowance for loan and lease losses (ALLL) preparation, inform a rigorous stress test and later be adapted for quantitative analysis to establish the allowance for credit losses (ACL) under the FASB’s current expected credit loss (CECL) notion.
Even for institutions with low or no history of credit loss, we can identify a segment-specific national and regional cohort of financial institutions to serve as a comparable baseline, as well as identify and support specific peers for inclusion. Emphasis is placed on a consistent process, which has proven to be an improvement in best practices commonly used in the industry to date.
When applying these models in a stress test application, we verify segment retention/growth scenarios and project the portfolio segments for two years using baseline, adverse, and severely adverse inputs to estimate credit losses over the scenario period. The intended use of this exercise is to inform input into a capital stress model or “break-the-bank” scenario.
When an institution’s allowance model is not useful for estimating losses due to economic uncertainty – commonly due to a lack of meaningful loss history – a segment-specific peer cohort can be used to substantiate qualitative adjustments under an ILM (current ALLL) framework. In contrast to a forward-looking CECL-notion ACL, model inputs would be held constant to present-period observable levels (known or knowable), and losses aggregated over an “incurred” (typically one year, but could vary with the use of a loss emergence period notion) projection period.
The models developed are appropriate for use under the CECL standard by applying a forecast and extending the time horizon to the life of the loan. A pro forma calculation using these models will communicate the likely impact of moving to the new standard and demonstrate preparedness for the transition.
About the Author
Garver Moore brings a decade of enterprise software, analytic, and advisory experience to Abrigo’s advisory team. Prior to joining Abrigo, Garver was a Technical Consultant with Accenture, and he later worked with C-suite executives on technology strategy and delivery as a Managing Partner of the Orange Advisory Group. Today, as the Managing Director of the Advisory Services Group, Garver collaborates with internal product specialists on our market offerings and helps clients develop strategies to better navigate federal regulations and optimize their institutions for growth. Garver earned his bachelor’s degree in electrical and computer engineering from Duke University.