With a little less than three years until private banks and credit unions must comply with the current expected credit loss (CECL) standard, making strides toward implementation may fall to the bottom of a 2020 priority list. As we move along in 2020, large SEC filers are already beginning to disclose the impact that CECL has had on their financial statements. When asked what advice they would give to their peers who have the 2023 implementation deadline, a common answer is that they wish they would have started earlier.
In a recent webinar with over 350 registrants from community financial institutions, Abrigo Consultant Zach Englert and Zach Langley, an Analyst at Abrigo, shared best practices they have learned from assisting SEC filers with their CECL models. In doing so, they explained how financial institutions with a 2023 deadline could emulate the successes of SEC filers while also preemptively avoiding their mistakes.
Data: The Foundation for CECL
“For our SEC filers who got a head start in 2017 – the ones who gave themselves a three-year cushion – when they saw their early preparation come in handy was when we started talking about data cleanup,” noted Langley. “As we went through this process with them, we identified the gaps that they had. It takes time to dig into data. It takes time, people, and a lot of resources to identify gaps, figure out how to bridge them, and do it. Doing that under a time crunch can make the situation more stressful. More mistakes can be made.”
On the topic of data, both Englert and Langley described how they had numerous situations where they thought that they were helping a public financial institution with their data under one set of assumptions and learned later that those assumptions were incorrect. Many institutions find that they may not have the right loan-level history, data fields, or meaningful losses at the start of their CECL prep. Because of this, it may be appealing to use third-party or peer data to fill in the gaps. Since data is the foundation for a defensible CECL model, if using either internal or external historical data to make an informed decision on your model, it is important to make sure that it is both verifiable and accurate.
“Data is so paramount in CECL. It makes sense that your model can only operate in the framework that your data creates. Whenever we’re looking at data, it’s often the ‘is this right’ scenario that institutions find to be the most challenging,” explained Englert.
CECL Implications/Economic Conditions
Overall, the move to CECL should improve consistency across the risk management processes. SEC filers who have applied the results of their CECL calculations to other functions of their business have discovered the most operational efficiencies. Having multiple risk areas operate under similar sets of assumptions and influence each other can save time by reducing duplicate data entry and give a more accurate picture of the overall health of the portfolio.
For example, the Bank of San Antonio leveraged connected inputs for both its allowance for credit losses and stress testing models. Because that information had been captured through its stress testing efforts, the bank didn’t need to input additional data on debt service coverage for CECL.
“Your allowance does have a material effect on financial reporting, and it should be included in your asset/liability forecasting,” said Englert. “All of this will play a role in the loan pricing conversation, capital planning, and stress testing.” In an ideal scenario, Englert explained, a bank would be actively assessing the effects of economic conditions or other credit risk factors on its portfolio. If the bank is seeing material impacts, then it should leverage that information for other elements in forecasting. “You need to be able to recognize the parallels,” Englert cautioned.
Getting a Head Start
Both speakers emphasized how CECL adoption is just a starting point, and they foresee changes being made to the standard as we move down the road from 2020 to 2023. Consequently, they recommend institutions start their CECL preparations as early as possible to be prepared for whatever roadblocks may arise. Additionally, having four quarters of parallel CECL calculations running against current accounting for credit losses under the incurred loss model can make the time after adoption easier. This will provide enough time for a financial institution to identify and fix mistakes, and ultimately get the number that they want. Getting a head start enables institutions to control their model instead of letting it control them.
“CECL isn’t something that you just get to wash your hands of when you hit your adoption date. That’s why having the chance to test your model before you go live with it is going to make for much easier maintenance. There’s a lot of long-term value in tackling these items as early as possible,” said Langley.
No matter what decisions your financial institution makes when preparing for CECL, or how long it takes to build out your model, you need to be able to document and support every decision you make. Auditors and examiners won’t just be looking to see that you’ve done the work and have a number, but you understand the assumptions, inputs, calculations, outputs, and resulting impact. One way to make your model stronger is to backtest your results, which can be conducted via a discounted cash flow or other methodology options.
If you are looking to make CECL progress throughout the remainder of the year but don’t know where to begin, Abrigo analysts can make a pro forma calculation of the standard’s impact at adoption for your institution. If interested, you can let us know here.
About the Author
Amanda Rousseau is a Segment Marketing Manager at Abrigo.