Jeff Judy

Jeff's Thoughts - June 15, 2016

CECL Arrives

In banking and credit risk management, the future just became more rigorous!

In late April, FASB voted to adopt their long-discussed standard for "current expected credit loss", CECL for short. And I cannot emphasize too strongly that the time to start working to meet this standard is right now.

It isn't that the deadlines for meeting this standard are unreasonable. Most community banks will have several years before they have to meet the CECL standard.

My concern is that the data required to determine "current expected credit loss" are not readily available at most institutions. It is all too easy to seriously underestimate the time and effort required to assemble all the information this standard will demand.

It will take some time just to begin the process of meeting the CECL standard. Each institution has to settle on a computational model for CECL, and then determine the information needed to feed that model.

Some of that data may not currently be collected. Or it may be gathered in some way, but perhaps not at the level of detail, or with the consistency, that is required. And it is very likely that to assemble the complete data, the bank will have to mine different systems, systems that may use different formats and that may not "talk" to each other well.

Start now. Understand the data requirements and, just as important, understand the challenges your own current data systems will present. Expect to do way more reformatting and cleanup of the data than you think will be necessary.

And at the same time, start working toward the philosophical change that comes with CECL. This isn't a simple change of terminology, where "expected loss" replaces "probable loss" or "accrued loss" as the "term of the day". This is much more than a cosmetic shift in how we talk about credit. A particularly challenging philosophical shift will require coming to grips with the impact of booking a new loan. No matter how pristine it is, it will have an impact on a CECL-based ALLL.

You are not only going to have to take a more forward-looking approach to losses, you are going to have to be able to show regulators that your model is forward-looking. Gone are the days when ALLL is determined by taking the last quarter's number and tweaking it up or down, based on perceived trends. For a long time, allocating ALLL has been as much art as science, but there will have to be a lot more science, so to speak, in ALLL allocations under the CECL philosophy.

In essence, ALLL protects against expected loss. But expected loss will, as a calculation, be held to a much more rigorous standard. And in turn, that calculation will depend on collecting data, charting the migration of credits among risk pools, and additional analysis that is currently well beyond what many institutions are doing.

Far more than a change in the way we talk about credit, in the banks that handle this new standard well, CECL will provide the impetus for a cultural shift. It will give banks a finer understanding of the health of their portfolios at the same time it forces a greater focus on the most probable future of those portfolios. Philosophically, I believe that's a healthy and helpful perspective for our industry.

Sure, it is a challenge. But it is one that smart banks will take up quickly and earnestly. The demands CECL makes on technology, operations, philosophy and culture suggest that foot-dragging could be disastrous.

We're all going to have to live with CECL in the fairly near future. Banks that embrace the new standard may not only avoid some painful discussions with regulators in a few years, they may gain powerful knowledge to enhance their own credit risk management and to avoid ugly surprises.