Key Takeaway: Under CECL, banks must estimate expected credit losses using forecasts that are "reasonable and supportable." Forecasts should extend as far as reliable data and assumptions allow, after which results must revert to historical averages — ensuring provisions are defensible, transparent, and aligned with regulatory expectations.
The Current Expected Credit Loss (CECL) standard requires banks to consider future events when assessing potential losses over the lifetime of their loans, rather than relying solely on historical data. For smaller community banks, compliance has proved challenging — many lack the resources of larger institutions, which would otherwise simplify modeling, data collection, and management.
What Does 'Reasonable and Supportable' Mean in CECL?
Estimated credit losses are considered reasonable and supportable when they are based on assumptions backed by documented data — applicable to loans, leases, and held-to-maturity debt securities. Institutions must not base estimates on general market sentiment or intuition; regulators require that any estimates or adjustments be grounded in data. Unemployment rates, shifts in borrower repayment patterns, and changes in property values are all valid examples. For smaller banks with limited loan histories, even small shifts in assumptions can produce dramatically different outcomes, making documentation especially critical.
How Long Should a CECL Forecast Period Be?
There is no mandated forecast length. The appropriate period depends on each bank's loan composition, available historical data, and modeling resources. Smaller community banks may only be able to forecast several quarters ahead based on local economic factors, while larger national banks can extend further using advanced macroeconomic modeling tools. The core expectation is that projections are defensible and grounded in reliable data — not that they predict the future perfectly.
CECL Reversion Methods
When forecasts can no longer be considered reasonable and supportable, banks must revert to historical averages. Common methods include:
-
Straight-line reversion: A gradual transition, well-suited for banks with stable portfolios.
-
Immediate reversion: An instant shift back to historical averages — the simplest and most defensible option for smaller institutions.
-
Stepped reversion: Incremental adjustments over time on a monthly, quarterly, or annual basis.
-
Hybrid approaches: Combinations of the above, suited for more complex, multi-risk portfolios.
Best Practices
Strong CECL forecasting requires consistency in methodology and documentation. Models should be built on institution-specific drivers — delinquencies, payment behavior, days past due — rather than industry-wide averages. Trusted macroeconomic data from sources like the FDIC and Federal Reserve can further strengthen credibility. Stress testing against rate shocks and deposit volatility helps validate reserve adequacy, and syncing CECL cycles with ALM improves cross-functional consistency.
How Empyrean Helps
Empyrean's CECL solution allows banks to run multiple scenarios while documenting each assumption — eliminating the need for external consultants. Integration with ALM, Budgeting & Planning, and Profitability modules ensures consistent data and assumptions across loan growth, capital planning, and provision forecasting, giving finance teams a single source of truth and a more complete view of balance sheet risk.
Want to learn more? Visit the Empyrean CECL page.