Articles & E-Books


Investigating CECL Methodologies

Nov 01, 2016

In our September 2016 e-Newsletter, we provided an overview of the new Accounting Standards Update (ASU) No. 2016-13, Measurement of Credit Losses on Financial Instruments. This standard will be effective in 2020 or 2021, depending on whether an institution meets the definition of a public business entity. Please refer to the article, Measuring Credit Impairment of Financial Instruments, for an overview of ASU No. 2016-13.

The Financial Accounting Standards Board (FASB) and regulators have emphasized that the standard is scalable for different institutions and that various methods are permitted to estimate life-of-loan losses. In other words, not every institution will implement the same methodology to estimate loan losses. This will allow institutions to implement a Current Expected Credit Loss (CECL) methodology that is appropriate for the size and complexity of their specific institution and loan portfolio.

Many believe CECL will be the most significant change to financial institution accounting in at least a generation. Institutions should understand that preparing to implement CECL will not be a simple undertaking. It may not be as difficult or time-consuming for community institutions as it will be for larger, more complicated institutions, but it will be a significant change, nonetheless, requiring time and substantial planning.

Investigating CECL Methodologies

In our last article, we suggested six steps institutions should consider now to prepare for CECL. One of those steps is to begin collecting data. Some data, such as loan-specific dates and amounts of originations and charge-offs, will be required by virtually any methodology selected to estimate loan losses, and institutions should begin collecting such data today. Other data will be specific to certain methodologies; consequently, institutions will have to select a methodology before they can be sure they are collecting all necessary data.

Since the selection of a CECL methodology will significantly affect the data to be collected, it is vital that financial institutions select the method(s) they plan to use sooner than later. Regulators have suggested methodologies should be selected by the end of calendar year 2017, and we believe this would be an appropriate milestone to target.

To help institutions investigate different CECL methodologies, the following provides a brief overview of some acceptable methodologies discussed in more detail in ASU No. 2016-13.

Cumulative Loss Rate Method

The cumulative loss rate methodology is used to develop a cumulative historical loss rate by identifying losses realized on loans outstanding as of a historical reporting date that generally coincides with the term of the loan pool. For instance, a December 31, 2016, analysis of a loan pool with a contractual life of five years might use charge-off history during the previous five years and the outstanding loan balance at December 31, 2011. The cumulative loss rate method is the simplest means to develop a quantitative base estimate of lifetime losses and could be done internally by many institutions. Unfortunately, this method assumes the risk characteristics of the loan portfolio today are consistent with the risk characteristics as of the reporting date used in the analysis. As a result, management will have to spend more time considering and supporting appropriate qualitative adjustments to the base estimate to adjust for current and forecasted changes in those risk characteristics.

Vintage Analysis

Vintage analysis looks at loan origination dates and loss rate patterns over time to estimate when loan losses will occur. To use this methodology, institutions will have to maintain more information than the cumulative loss rate method requires, including loan origination dates and amounts and loan charge-off (less recovery) dates and amounts. From this information, institutions can create vintage tables to analyze loss rate patterns and develop estimated loan losses. Although more involved than the cumulative loss rate method, some institutions may still be able to complete this analysis internally without the use of specialized software. This method has the added benefit of inherently developing support for some of the qualitative factors management will consider.

Other Methodologies

Other acceptable methodologies include migration analysis, probability of default, and discounted cash flows. Each of these methods yields more precise quantitative estimates that can be calibrated without relying on significant qualitative factors. Migration and probability of default analyses are much more sophisticated than the previous methods discussed and may require the use of thirdparty software. Discounted cash flow analysis can be a relatively easy method for estimating loan losses for an individual loan but would require significant computational power to use for a pool of loans.

Concluding Thoughts

Although it may seem that there is plenty of time to implement CECL, the new accounting standard will require significantly more data over longer periods of time. To identify the required data, institutions need to be thinking about the method(s) they plan to use for CECL. Institutions should target December 2017 to select appropriate methodologies so they have adequate time to collect necessary data for the applicable methodologies. For more information on methodologies that are available to financial institutions, please consider attending Wipfli’s upcoming webinar, Getting Ready for CECL, on December 1, 2016. If you have questions today, please contact your Wipfli relationship executive or email us at


Brett D. Schwantes, CPA
View Profile