March 5, 2018

# CECL Models – Loss Rate Analysis

By Brenda DeCosta, Partner, Assurance Services

On June 16, 2016, the Financial Accounting Standards Board (“FASB”) issued ASU 2016-13, Financial Instruments – Credit Losses (Topic 326), Measurement of Credit Losses on Financial Instruments. This standard is expected to significantly change the method of calculating the allowance for loan losses by requiring the use of the Current Expected Credit Losses (“CECL”) Model.

As financial institutions continue to focus on the implementation of the CECL, the persistent question is what will the calculation actually look like? After attending the Banking CPAs Summer Conference in Nashville, we have compiled some examples of the CECL methodologies as a visual aid. Be advised that these examples are extremely simplified in order to properly demonstrate the theory behind each methodology (actual results may vary). What is considered to be the simplest form of the calculation — the loss rate analysis — follows.

For background on the following example, please note:

1. This is a very simplified approach.
2. It is based on a portfolio of 10-year amortizing loans.
3. Said loans were originated over the last 10 years.
4. Current balance is \$3,000,000.

 Year Portfolio Balance Annual Losses Annual Loss Rate 2010 Loan Losses 2010 \$2,000,000 \$7,000 .35% \$0 2011 \$2,100,000 \$7,350 .35% \$7,000 2012 \$2,200,000 \$7,700 .35% \$5,800 2013 \$2,300,000 \$8,050 .35% \$4,900 2014 \$2,400,000 \$8,400 .35% \$4,100 2015 \$2,500,000 \$8,750 .35% \$2,900 2016 \$2,600,000 \$9,100 .35% \$2,300 2017 \$2,700,000 \$9,450 .35% \$1,800 2018 \$2,800,000 \$9,800 .35% \$1,200 2019 \$2,900,000 \$10,150 .35% \$0 2020 \$3,000,000 \$10,500 .35% \$0

The annual loss rate (i.e., the annual losses divided by the portfolio balance) was 35 basis points each year within the example for simplification. The calculation is then prepared by summarizing the cumulative losses on the \$2,000,000 of 2010 originated loans over their 10-year term (i.e., \$30,000 as highlighted below).

 Year Portfolio Balance Annual Losses Annual Loss Rate 2010 Loan Losses 2010 \$2,000,000 \$7,000 .35% \$0 2011 \$2,100,000 \$7,350 .35% \$7,000 2012 \$2,200,000 \$7,700 .35% \$5,800 2013 \$2,300,000 \$8,050 .35% \$4,900 2014 \$2,400,000 \$8,400 .35% \$4,100 2015 \$2,500,000 \$8,750 .35% \$2,900 2016 \$2,600,000 \$9,100 .35% \$2,300 2017 \$2,700,000 \$9,450 .35% \$1,800 2018 \$2,800,000 \$9,800 .35% \$1,200 2019 \$2,900,000 \$10,150 .35% \$0 2020 \$3,000,000 \$10,500 .35% \$0 \$30,000

Additionally, prepayments are automatically considered, as a study on the history of the loan portfolio would need to be completed to incorporate that into the average term (in this case, 10 years). Be advised that the historical loss rate does not take the 35 basis point annual loss rate and multiple it by the 10-year term, which has been a common misconception. Instead, the lifetime historical loss rate is calculated by taking the \$30,000 in cumulative 2010 loan losses and dividing it by the \$2,000,000 in 10-year amortizing loans originated in 2010, for a loss rate factor of 1.50%. This demonstrates the difference between the current historical loss rate calculations, based on a period of time, versus the CECL concept of the lifetime loss rate.

Your institution would then need to factor in the qualitative factors (i.e., change in real estate values, unemployment, loan underwriting, etc.) and the new concept of a forward-looking component under CECL. The issue with the qualitative factors and forward-looking component is the need for detailed and supportable factors. For purposes of this example, let’s assume the Q factors for this 10-year amortizing loan portfolio totaled 15 basis points, for a total CECL rate of 1.65% on a \$3,000,000 loan portfolio, resulting in a required allowance for loan and lease losses (“ALLL”) of \$49,500.

The data requirements for the loss rate analysis are not overly cumbersome, are more than likely readily available to your institution, and include the following:

1. Total loan principal as of a date corresponding to the term of the loan pool.
2. Loan losses (recoveries) by date/fiscal year.
3. Origination date of loans with losses (recoveries).

Overall, there are both pros and cons to this loss rate calculation methodology. The pros include that it is the simplest method and requires the least amount of additional/new data. The cons are that your institution:

1. May need to disaggregate pools by maturity.
2. Needs to carefully consider changes to the risk characteristics of the portfolio.
3. Needs greater analysis for the Q factors.
4. May likely end up with a higher ALLL than other, more comprehensive methods.

• Assurance
• Boston, MA