Four Steps to Optimize the Impact of CECL on Profitability
Data, data, data. You have it, but now what? Financial institutions are being bombarded with the need for data in order to be in compliance once the Allowance for Loan and Lease Losses (ALLL) regulatory changes using the Current Expected Credit Loss (CECL) model comes into effect.
If I have all the data in the world, and know not what to do with it, what good is it? The service providers are all gearing up to store massive amounts of data for years on end, and it is extremely important to do so now while leading into the implementation dates. However, clearly thought out and well-defined analytics is where the focus should be. This cry is not to downplay the need for the data, but to redirect the attention towards another aspect of the upcoming changes: the economic drivers of the calculations.
Step 1: Segment for Impact and Risk
What are the building blocks? Initially, start with segmenting/pooling the balance sheet, becoming more refined while grouping like-kind products and services together. Real Estate Loans, Commercial Loans, Individual Loans, and Other Loans & Leases all have descriptive differences, but these categories don’t hone in on the critical economic drivers. The subgroups such as Commercial & Industrial, Construction & Development, 1-4 Residential, and Home Equity Loans, which you find in the Uniform Bank Performance Report (UBPR), are more granular but take it a step further by defining fixed, floating, and adjustable by their relevant indexes within a manageable population. The economic drivers of those indexes can be easier to define and can have differing influences when broken down.
Step 2: Use History to Prepare for the Future
Where do you start after you’ve defined the segments? Start with the historical loss data that has been the basis of the allocations in the past. Then choose a CECL methodology type to make the adjustments. Some options available include migration analysis; vintage analysis; discounted cash flow; or the probability of default method (PD/LGD/EAD).
Step 3: Identify Economic Drivers
Next step is to identify which economic drivers impact these categories, top to bottom. This will not be an easy process. Some in-depth thought needs to be used in isolating these drivers. Some can be universal for example, the treasury rates or advance rates. But real gross domestic product, nominal gross domestic product, the unemployment rate, nominal disposable personal income might have a breadth that is too wide, and—depending on the institution’s footprint—using regional indicators might be more indicative.
Step 4: Learn and Repeat
Keep in mind, you may not find the right combination on the first try. Consider a variety of metrics and begin to test the impact on profitability. Find what matters and identify what economic data helps to explain the institution’s loss patterns. Use those loss correlations in defining the basis of the future adjustments. Document the findings for both the correlations used in the assessment and those that did not.
Yes, gathering data is extremely important to do. Do so now while leading into the implementation dates but most importantly, design a systematic process for evaluating your portfolio that can be repeatable, directionally consistent, reasonable, and supportable.
Posted on Saturday, October 13, 2018 at 1:45 PM
by John Robertson