Skip to main content

Quantitative Methodologies for CECL

April 27, 2016
Read Time: 0 min

by Chris Emery, MST Special Projects Director

Much of the publicity around FASB’s CECL proposal centers on the concept of adjusting loss rates for “reasonable and supportable forecasts.” There’s been a lot of talk about what economic forecasts should be used, the reliability of said forecasts, how far into the future an institution should look, etc. While these are certainly issues that need to be discussed, there is risk of ignoring what will also be a potentially significant change under CECL: what is required for the qualitative portion of your allowance. After all, the basic premise of CECL is to base an allowance on information about “past events, current conditions, and reasonable and supportable forecasts.”

Perhaps some of the reason the historical qualitative portion of CECL hasn’t gotten as much buzz is because financial institutions feel like they have a better handle on this part of the guidance. After all, they have been calculating some historical qualitative portion under ASC 450-20 (and previously FAS 5) for years. However, there are some new challenges that will come along with this part of the allowance for CECL that need to be considered.

The first challenge is that most banks are currently relying on annualized net charge-off rates as the basis for their ASC 450-20 reserve. The allowance generated based on these rates would then theoretically represent the losses expected over the next 12 months before any qualitative adjustments. However, under CECL, allowance will need to be based on the lifetime expected losses of the financial instruments, not simply the next 12 months. FASB has also made it clear that simply multiplying an annual loss rate by the average life of a loan in that segment would also not represent an appropriate number. Instead, what will need to be calculated is the historical cumulative lifetime loss rate for a particular segment of loans. In theory, this would represent all of the losses that have occurred over the lifetimes of loans in the segment divided by the totals balances of the loans in that segment.

FASB’s CECL guidance gives several examples for types of quantitative analysis in their draft document, all of which are types of cohort analysis. These include Estimating Expected Credit Losses Using a Loss Rate Approach, Estimating Expected Credit Losses on a Vintage-Year Basis, and Estimating Expected Credit Losses Using an Aging Schedule. In cohort analysis, loans of similar characteristics (or “cohorts”) are tracked over time. In a cohort analysis, there are two key capabilities required by the qualitative model: 1) tracking what segment loans exist in over time and 2) tracking what losses occur over the lifetimes of these loans. The only key difference in the three methods FASB describe is what the segments are based on. In the first example, they are purely based on loan types, in the second example, loan types and vintage years, and in the third example, loan types and delinquency rates. Different segmentations may be more or less appropriate for different types of loans or different financial institutions.

By tracking the populations of these loans over time, you can begin to see the long-term historical cumulative lifetime loss rate for a given segment. This type of analysis also inherently takes life of loan and prepayment calculations into account, so these items would not necessarily have to be adjusted for separately, provided the financial institution expects loans to continue to be originated and pay off at similar rates going forward as they did in the historical averages.

For some existing MST clients using the Loan Loss Analyzer, the cohort analysis example may sound very familiar to you, and it should! The type of cohort analysis FASB is describing is very much like our existing Migration Analysis methodology. For clients already using this type of analysis, there may be only minor adjustments necessary. These may include “uncapping” the loss horizons so that losses that occur will migrate back throughout the life of the loan and potentially also adding additional segmentation criteria to some existing pools. For those MST clients who are not already using any type of Migration Analysis methodology, there is still good news! Every quarter that you run your analysis in the LLA, you are storing the type of loan-level analysis data that would be required for cohort or Migration Analysis once you are ready to implement.

Hopefully this sheds a little light on some of the types of qualitative methodologies that will be allowable or appropriate under CECL. Some of this is obviously still subject to change since we don’t yet have a final standard, but we would not expect any final standard to deviate from the principles we’ve described here, as most if not all of them have been consistent throughout the exposure draft and subsequent board decisions.

About Abrigo

Abrigo enables U.S. financial institutions to support their communities through technology that fights financial crime, grows loans and deposits, and optimizes risk. Abrigo's platform centralizes the institution's data, creates a digital user experience, ensures compliance, and delivers efficiency for scale and profitable growth.

Make Big Things Happen.


Looking for Banker’s Toolbox? You are in the Right Place!

Banker’s Toolbox is now Abrigo, giving you a single source for all your enterprise risk management needs. Use the login button here, or the link in the top navigation, to log in to Banker’s Toolbox Community Online.

Make yourself at home!