Banks are beginning to release the impacts of COVID-19 on their credit loss reserves for the first quarter of 2020, and, unfortunately, it isn’t pretty. In fact, it’s downright UGLY. Such reserves, held by banks to be used in the event of loans defaulting, are reassessed each reporting period (e.g., quarterly by most banks). But why such a drastic increase, and so quickly? Doesn’t the current expected credit losses (CECL) model under ASC 326 Financial Instruments – Credit Losses require entities to forecast expected losses such as this? Let’s dive a little deeper to explore the concept of a “forecast” under the standard and then look at the impacts COVID-19 has had on credit loss reserves.
The CECL model, outlined in ASC 326, requires lifetime expected credit losses be recognized immediately when a financial asset is originated or purchased. So, what does this mean in the context of a loan on the bank’s books? When a bank originates a loan, on Day 1, the bank is required to recognize the amount of losses it estimates it will not collect over the period the loan remains held and outstanding. The estimate of losses is recognized by using an allowance for credit losses (ACL) contra-asset account to reflect the amount the entity does not expect to collect of the amortized cost of the asset, in this case, a loan, during the asset’s contractual life. The ACL estimate is to be reassessed each reporting period and is subject to reversal as conditions change, whether favorable
When developing an estimate of expected credit losses…an entity shall consider available information relevant to assessing the collectability of cash flows. This information may include internal information, external information, or a combination of both relating to past events, current conditions, and reasonable and supportable forecasts.” ASC 326-20-30-7
While the standard does not prescribe a required methodology to be used when forming an estimate of the ACL, it does require an entity to consider past events, current conditions, and reasonable and supportable forecasts in the estimate.
Reasonable and supportable forecasts
How is the “reasonableness” of a forecast determined? How is a forecast deemed “supportable”? The FASB left us in the dark with these two terms, as they are not defined in the standard. Entities must utilize significant judgment when determining forecasts of future conditions. Documentation, transparency, and consistency are extremely important when it comes to reasonable and supportable forecasts. One source for determining the reasonableness of the estimate might be the directional consistency of historical information driving the logic for your forecasts.
Will entities need to use a crystal ball? Time machine? Some other method of divination?
The CECL methodology does not mean banks will need to predict the future to make their estimates each period; rather, banks should use the information available to make a reasonable and supportable estimate. History is often a great place to start when trying to forecast the future – the FASB even thinks so! In ASC 326-20-30-8, the FASB explained “historical credit loss experience…generally provides a basis for an entity’s assessment of expected credit losses.” However, the FASB does note that “an entity shall not rely solely on past events to estimate expected credit losses. When an entity uses historical loss information, it shall consider the need to adjust historical information to reflect the extent to which management expects current conditions and reasonable and supportable forecasts to differ from the conditions that existed for the period over which historical information was evaluated.” (ASC 326-20-30-9)
Entities could consider economic relationships found in their asset portfolios to assist with their forecasts. Examples for a bank include:
- Is the borrower’s industry contracting, stable, or thriving? A thriving industry in the borrower’s market may indicate continued success or it may indicate the market will soon reach a plateau.
- Are there local, regional, or national data points that can be used to assist in predictions of the economic life cycle of the borrower’s business? Examples may include unemployment rates, interest rates, foreign exchange rates, etc.
- Are there other external factors that could impact the borrower’s repayment of the loan? If the bank has a large agricultural lending portfolio, perhaps weather trends should be considered.
As you can see there are numerous data points that an entity may consider when forming their forecasts. Information sources for these data points can include publicly available external forecasts, internal experts, and external experts. There are many acceptable data sources for historical loss information and entities are able to use any combination of these sources. For example, an entity may have sufficient historical losses for some products but a new product may require the entity to utilize peer loss data.
Do entities need to search the ends of the Earth for information?
Does this mean entities need to use sophisticated data gathering and analysis techniques to form their forecasts? Nope! While banks are expected to collect data and consider past historical information, present data, and future conditions when applying the CECL model, there is no required methodology or level of complexity prescribed by the standard. Entities are certainly NOT expected to search all possible information not reasonably available without undue cost and effort or create a hypothetical pool of assets. In fact, the guidance notes internal information may be sufficient! Each institution’s approach will be unique and the definition of reasonable and supportable may vary. Complexity levels also will vary based on the complexity of the institution itself. What an institution deems to be “undue cost and effort” this quarter, could change next quarter, as more and more data becomes available.
Speaking of more data becoming available, at the beginning of this post, I asked why COVID-19 impacts weren’t included in the original estimates made by entities when adopting ASC 326 as of January 1, 2020. Based on what we’ve discussed thus far regarding forecasts, we now have an answer to that question! While some may argue an entity should have considered a worldwide pandemic in a future forecast of losses formulated upon adoption of ASC 326, it is extremely unlikely that such a forecast would be considered BOTH reasonable and supportable. With each passing day more and more information becomes available related to the impacts of COVID-19 on both individuals and businesses. As the first quarter 2020 financial results are being released, we are seeing detailed discussions on how banks are considering the impact of COVID-19 on their ACL estimates.
Impacts by the numbers
As a reminder, ASC 326 and its CECL model was required to be implemented by certain larger, SEC filers on January 1, 2020. (See this post for details on effective dates and implementation delays.) GAAP Dynamics performed an analysis of the impacts of implementing ASC 326 on 20 banks here. The average impact for that sample was an increase in the reserves of 43%, or $1.3 billion of additional reserves added. We used a sample of 8 banks to illustrate the impacts of COVID-19 on the banks' reserve estimates.
In the chart below, we note that the implementation of ASC 326 resulted in an average increase of 39%, or $2 billion of additional reserves for the 8 institutions listed. That’s quite an increase! And that was BEFORE the impacts of COVID-19 were taken into account. After considering COVID-19, these 8 banks added an additional $2.9 billion, or 29%, to their reserves during the first quarter of 2020. While not all of this increase is directly attributable to COVID-19, the bulk of it would be!
This means that since December 31, 2019, as a result of the implementation of ASC 326 and the impacts of COVID-19, the credit loss reserves at these 8 banks have increased by a whopping $4.9 billion!
Unfortunately, this is only the first quarter that the impacts of COVID-19 have been incorporated into the forecasts used in ACL estimates. The ripple effects of the global pandemic will undoubtedly continue into the future. The implementation of ASC 326 was already considered a monumental task by banks, and now, they have another uncertain data point to incorporate into their forecasts, all while making sure these forecasts are reasonable and supportable.
Confused about CECL? Have a question on applying ASC 326 or IFRS 9? We currently have online courses on the CECL model as well as IFRS 9 across a variety of delivery methods. Need additional help? Let’s talk. We’re always glad to help!
About GAAP Dynamics
We’re a DIFFERENT type of accounting training firm. We don’t think of training as a “tick the box” exercise, but rather an opportunity to empower your people to help them make the right decisions at the right time. Whether it’s U.S. GAAP training, IFRS training, or audit training, we’ve helped thousands of professionals since 2001. Our clients include some of the largest accounting firms and companies in the world. As lifelong learners, we believe training is important. As CPAs, we believe great training is vital to doing your job well and maintaining the public trust. We want to help you understand complex accounting matters and we believe you deserve the best training in the world, regardless of whether you work for a large, multinational company or a small, regional accounting firm. We passionately create high-quality training that we would want to take. This means it is accurate, relevant, engaging, visually appealing, and fun. That’s our brand promise. Want to learn more about how GAAP Dynamics can help you? Let’s talk!
This post is published to spread the love of GAAP and provided for informational purposes only. Although we are CPAs and have made every effort to ensure the factual accuracy of the post as of the date it was published, we are not responsible for your ultimate compliance with accounting or auditing standards and you agree not to hold us responsible for such. In addition, we take no responsibility for updating old posts, but may do so from time to time.