The reserving process design should include:

• Seamless integration with data, disclosures and analytics;

• Flexibility to perform a variety of CECL-compliant methods within a controlled framework;

• The ability to apply expert judgment via post-processing adjustments; and

• The ability to identify and review assets evaluated on an individual basis based on borrower-specific facts, for example, where the practical expedient for collateral dependent assets is applied.

Because CECL expands the information considered in the estimate to include forecasted conditions, the factors driving the change in the reserve will differ from that of current U.S. GAAP. Consider this analogy – today, you pull your umbrella out when it is raining, but CECL will require you to pull your umbrella out when there is a forecast of rain! Therefore, it should come as no surprise that the CECL reserve will be more volatile than the allowance today. It is for these reasons that credit unions should leverage a proper reporting framework that links the changes in expected credit losses at the instrument level to all changes, including changes in forecasted conditions, credit risk and portfolio composition.

The reporting framework should include:

• Push-button reports for internal management packages and disclosures that are consistent each reporting period;

• True ad-hoc analytics drillable to the asset level in order to address impromptu questions from internal and external stakeholders;

• Attribution analysis and migration analysis enabling comparison between prior and current period reserves;

• Scenario analysis comparing results using different assumptions and forecasts; and

• Transparency into the reversion to historical loss information during the contractual periods beyond the reasonable and supportable forecasts.

Data is the beginning of all business processes and the common thread across functional areas of the reserve. Consolidate all data requirements across modeling and reporting, and identify the least common denominator. Application-ready data – data suitable to be processed by models and accessible for reporting and analytics – must be clean, valid and normalized. Finally, integrate data from the model inputs to the model results in order to automate flexible multi-dimensional reporting. The data framework should include:

• Input controls to ensure appropriate transmission of files and file integrity;

• Validations to ensure that necessary data elements are complete, correct and usable;

• Aggregation of instruments for collective evaluation and identification of assets to be evaluated on an individual basis; and

• Tracking and maintaining of asset records including the source, transformations and ultimate use (i.e., data lineage).

A successful CECL transition requires taking a holistic view of the end-to-end reserving process and leveraging the right tools.

Lauren Smith is the Director of Accounting Policy & Research at SS&C Primatics. She can be contacted at  [email protected].

John Lankenau is the SVP, Product & Operations at SS&C Primatics. He can be contacted at  [email protected].