What’s really at stake with your institution’s CECL data?

Since the Current Expected Credit Loss (CECL) standards were finalized in June 2016, one major area of concern among bank and credit union CECL teams and their CFOs has been that the new set of loan loss calculations will require extremely granular and high-quality data.

Part of the concern is based on what data has been captured and what data issues exist. The bigger concern, however, is that the new standards will almost certainly require financial institutions to increase their loan loss reserves – perhaps to a level as high as 2% or 2½% of assets, especially for loans with longer lives.

With this potentially large reserve increase in mind, the loan loss calculations will need to be as accurate as possible for an institution to avoid setting aside a higher than necessary reserve. Having loss estimates as accurate as possible will be based partly on the amount of institution-specific key data elements (KDEs) a bank or credit union has and the amount of general market data the institution needs to supplement its own KDEs.

The importance of KDEs is highlighted in a Situs white paper, CECL Data: Don’t Wait, Start Collecting Key Data Elements Now. The author of the paper is Jeff Prelle, managing director of analytics and head of risk modeling at MountainView Financial Solutions, a Situs company.

“For well-run institutions, using their own data should, at least potentially, lead to setting aside smaller reserves from a CECL perspective and increased production efficiency due to data-driven decision making,” Prelle said in the paper.

Prelle added that he believes the concern about the effect of significantly increased reserve levels upon CECL implementation is behind some of the answers CECL teams provided in MountainView’s CECL implementation progress surveys earlier this year. In answers to the credit union version of the survey, data was the greatest implementation challenge. In the bank version, data was the third-largest challenge (behind model development and model validation), but perhaps this ranking was due to many banks having CECL implementation dates earlier than credit unions and having already resolved some of their data problems.

Situs’ CECL KDEs white paper is available for download. The paper makes the case for exceptionally accurate KDEs and rigorous quality control on those KDEs.