CECL is a flexible standard that allows many methods for estimating an institution’s Allowance for Credit Losses (ACL). For participation portfolios, choose one that reflects the variety of loan products and originators you hold. Participations increase portfolio diversity, so the resolution of your CECL analysis has to sharpen accordingly.
Three WARM methods, three levels of granularity
All three use the Weighted Average Remaining Maturity (WARM) method. Ignoring qualitative adjustments, each segment’s ACL is its outstanding balance × an estimated annual loss rate × a WARM estimate in years, summed across segments. They differ only in how the loss rate is estimated:
| Method | Loss rate source |
|---|---|
| Portfolio level | One rate for all participations, from the owner’s losses in that portfolio over 3 years |
| Segment level | One rate per loan segment, from the owner’s losses in its own portfolio over 3 years |
| Segment/originator level | One rate per segment and originator, from each originator’s reported losses over 3 years |
Why granularity matters
First, different products and different originators produce different loss outcomes, so specific data gives more accurate estimates. Second, data availability: participation owners rarely have three full years of history with their current mix, and often buy loan types they rarely originate. Originators, by contrast, usually sell participations only after years of building a robust loss history.
What the comparison shows
Applied to four participation portfolios, the results varied. In two, segment-level reserves ran higher than segment/originator; in the other two, the reverse. The segment/originator method moved more month to month, reflecting real changes in portfolio composition. The portfolio-level method barely moved at all and, in three of four cases, recommended significantly smaller reserves than the other methods.
Best practice
Insufficiently granular CECL creates the risk of over- or under-reserving, and of sudden ACL jumps when lagged estimates finally catch up. Credit unions should use methods that reflect the variety among participation partners and assets. This approach is easy to automate, internally or with a technology provider.