How Do Credit Unions Ensure CECL Data Adequacy?
“When considering member business lending and the consumer portfolio where many credit unions focus, change management to prepare for CECL can be even more difficult with a credit union,” says Danny Sharman, risk management consultant at Abrigo during a recent CECL webinar. “The first thing credit unions can do to prepare is begin collecting data on multiple segmentations.”
Data adequacy tips
During the webinar, Sharman addresses the importance of understanding CECL guidance and the three methodologies for data collection. He also outlines the following points to keep in mind when collecting and storing data:
1. The data needs to be labeled meaningfully and consistently (e.g., to differentiate between loan, product, call or purpose code).
2. Data does not contain duplicate fields, rows or entities.
3. There are no inconsistencies in values (e.g., truncated by 000s vs. not truncated).
4. Data is stored in the right format – as numbers instead of text for calculations.
5. The file extracted from the core system is stored as the right file type and in the right, secure location.
6. File creation is automated instead of a manual process. This reduces risk of missed archives or overwriting.
7. Data is reliable and standardized throughout the institution. The institution can define and socialize a data dictionary that keeps all parties up to date.
8. Data fields are standardized and governed to ensure consistency.
9. Data storage does not have an archiving time limit (e.g., 13 months).
10. Data files are in an accessible format (not PDF).
11. Archiving function captures data points required to perform range of robust CECL methodologies.
“The challenge associated with credit unions is to ensure that the data is consistent across data sources. Having multiple sources is much more common in credit unions than community-sized banks,” notes Sharman. The problem can be exacerbated if different teams in the credit union have a different understanding of what is archived. For example consumer lending teams may need to track different risk rating criteria than a member business lending team. How are those expectations set and upheld?
How to fix data inadequacy
If found that the institution’s data capturing and storage solutions do not meet the more robust requirements of the proposed CECL model, there are three ways to correct the data.
1. Contact core provider & inquire about data archiving
While credit union data is already stored in its core and/or decisioning system(s), oftentimes these systems only store data for up to 13 months. In order to meet CECL requirements, contact the core provider regarding data archiving options.
2. Work with IT department to create an in-house solution
PDFs with data tables will not help the credit union achieve its reporting goals. When credit unions take on the responsibility for pulling together data archives, the IT department will most likely need to develop a query-able and reportable format for data storage.
3. Partner with a third-party, trusted vendor
If a credit union chooses to utilize a provider like Sageworks ALLL, the vendor will be responsible for data storage and management and can act as a bridge between the institution’s core data and an ALLL solution.
In a recent Abrigo poll, 35% of credit unions surveyed indicated introducing new models and executing various methodologies was the biggest challenge in planning for CECL implementation, while 30% cited data retrieval. Understanding methods for data gathering and storage, while outlining actionable steps is the key to preparing for CECL today.