Knowledge Center >

CECL 2022 Mid-Year Review

Introduction

As part of an annual study that began in 2020, Valuant conducts analysis on ASC 326, commonly referred to as “CECL”. The study contains key data statistics and insights around the US Regional and Community Banking sectors and the impact CECL has on their Allowance for Credit Loss (ACL) Coverage Ratios. The initial study included analysis of 137 banks that adopted ASC 326 as of 1/1/2020 and expanded to 173 institutions in the 2021 study. These banks included regulated financial institutions whose stock traded on a Major Public Exchange (NYSE, NASDAQ, OTC) and had total assets of less than $400 Billion as of the applicable measurement date. The analysis is evaluated in groupings based on both geographic region (defined by U.S. Census Bureau Region) and total asset size (aggregated into 4 categories – less than $10 Billion, between $10-$25 Billion, $25-$100 Billion, and greater than $100 Billion). We continue to use these characteristics in our study given both their relevance and for comparability to prior year studies.

 

This whitepaper will detail updates to key data points and analysis from the prior year studies through 6/30/2022 reporting periods. We will also discuss how the industry responded to credit loss reserving through the first half of calendar year 2022, our expectations for future reporting periods in 2022, and key items for 2023 adopters. It should be noted that most of the dollar volume of assets in the Financial Services Industry are already reporting under the CECL framework. However, these companies represent a minority in terms of the total count of institutions. All private financial services companies (including Credit Unions) and Smaller Reporting Companies (SRC) Public Companies (as defined by the U.S. Securities and Exchange Commission or “SEC”) will be required to comply with CECL beginning with fiscal year end dates after 1/1/2023.

 

Study Data Overview

Data was sourced from SEC Form 10-K and 10-Q to identify the most accurate values and allow for review of detailed commentary. While CECL applies to both Unfunded Commitments and Held to Maturity Securities, our study focuses on Loans Receivable. Furthermore, tax impacts are not considered within the study. It should also be noted that the ACL coverage ratios were calculated using disclosed book balances rather than amortized cost basis given their relative availability and comparability. As prescribed by CECL guidance, the amortized cost basis adjusts book balances for items such as deferred fees and costs, premiums or discounts, and accrued interest receivable (if the company elects to include it in amortized cost basis). As such, there may be slight variation in ACL coverage ratios disclosed in financial reporting than used in this study. 

As noted in the introduction, the 2021 study included 173 CECL adopters. The 2022 refresh includes 190 institutions in total, of which 175 are active as of 6/30/22. While the underlying data is updated as of 6/30/22, the institutions evaluated remains constant, updated for any new adopters and those that no longer exist due to acquisition. Note that there may be some variation in population counts compared to prior studies due to these factors. The chart and table below detail the composition of the dataset by region and asset size.

 
The following chart reflects the various adoption dates of the underlying population. Note that because some of the institutions included the study do not have traditional fiscal year ends, there are a few non-year end adoption dates.

Reporting Period Recap Q3 ’21 – Q2 ’22

Wrapping up the last two quarters of 2021, zero banks in the dataset elected to early adopt in the latter half of the year. However, institutions picked up speed as we moved into the new year. As of Q1 2022, 12 institutions entered the study; about half of which related to those banks who had further delayed adoption due to the Consolidation Appropriations Act. Including these institutions, the overall reserve ratios dropped from an average of 1.23% at 12/31/21 to 1.18% as of 3/31/22. The following chart displays these coverage ratios by region and by asset size.

Similarly, we continued to see a drop in the average reserve ratios to 1.16% in Q2 ’22. Over the last year we’ve observed a steady decline in reserves that were built up in part to the COVID-19 pandemic. At this point, the period over period unwinding is beginning to taper off. As concerns have shifted over the last few quarters due to higher inflation, geopolitical unrest, and anticipated increases in unemployment rate, many institutions are monitoring growing economic uncertainty. As CECL guidance requires banks to leverage a reasonable and supportable forecast in calculating expected losses, commentary from the Q2 ‘22 10-Qs suggest that many institutions have increased or are considering increasing reserves due to the current economic environment. We examined disclosures for the banks with the largest changes in allowance from Q1 ‘22 to Q2 ‘22 to attempt to isolate significant factors attributed to the change in reserves. A few pieces of specific commentary below:

 

SVB Financial Group, 6/30/22 10-Q

The provision for loans…for the three months ended June 30, 2022 was driven primarily by a deterioration in projected economic conditions. We assigned a higher weighting to our downturn outlook scenario to reflect our best estimate of those forecasts [1].

Citizens Financial Group, 6/30/22 10-Q

Our June 2022 qualitative consideration for macroeconomic risk reflects the Federal Reserve’s aggressively tightening monetary policy and the contraction of fiscal policy, as pandemic-related support winds down. These conditions, weighed together with the impacts of Russia’s invasion of Ukraine on key global commodity prices, labor shortage-related wage increases and continuing supply-chain challenges contributing to surging inflation, possibly may push the U.S. economy into a mild recession and create volatility in key macroeconomic variables, including GDP and employment [2].

Northwest Bancshares, Inc, 6/30/22 10-Q

The current period provision was driven by loan portfolio growth and the slower economic growth forecasts. The lack of provision in the prior year was driven by the improvements in the economic forecasts compared to the uncertainty that existed in 2020 to the industries impacted by COVID-19…In determining the amount of the current period provision, we considered current and forecasted economic conditions, including but not limited unemployment levels, expected economic growth, bankruptcy filings, and changes in real estate values and the impact of these factors on the quality of our loan portfolio and historical loss experience [3].

 

As highlighted in the disclosures, banks will need to continue assessing current economic conditions and adjust model inputs and assumptions on an ongoing basis as required under CECL guidance. Commentary from footnotes also highlights the importance of supplementing with qualitative factors where necessary to capture additional economic risks that may not yet be contemplated in mechanical forecasts.


Analysis by Region, Asset Size and Loan Composition

As previously mentioned, we continue to analyze the dataset using segments based on geographic region (defined by U.S. Census Bureau Region) and total asset size (aggregated into 4 categories – less than $10 Billion, between $10-$25 Billion, $25-$100 Billion, and greater than $100 Billion). In addition to those two breakouts, new to the 2022 refresh we have also sub-segmented the data based on loan composition (Commercial, Commercial Real Estate, Residential Real Estate and Other Consumer loans).

In terms of asset bands, the coverage ratios for Q3 ’21 – Q2 ’22 remain consistent with the trends we’ve seen over the past year. Although the larger asset band, institutions greater than $100B, was by far the most reactive to the economic distress and uncertainty that arose out of the COVID-19 pandemic, these institutions have seemingly unwound reserves getting them back at or near pre-pandemic levels. Likewise, the $25-$100B band has also returned to near pre-pandemic reserve levels. Interestingly, the two smaller asset groups, under $10B and $10-$25B, have been much slower to return back to the reserve levels seen at adoption. Specifically for this quarter, reserves decreased at a much smaller magnitude than they had previously since Q4 ’20 indicating that these institutions may have lasting concerns around the economic environment and may not be comfortable returning to pre-pandemic reserve levels quite yet.
 
Next looking to geographic regions, trends in average coverage ratio also aligns to that which we’ve seen across asset bands. As mentioned previously and further supported by the trend chart above, over the past year we have continued to see institutions all across the country steadily release reserves that were built as part of COVID-19 pandemic. While these reductions in reserve have seemingly tapered off this quarter, none of the regions have returned to the reserve ratios that existed at adoption. Mirroring the trends we’ve observed across asset bands, this is likely tied to growing economic concerns and a lack of comfortability in returning to pre-pandemic reserve levels. As evident in the trend chart, the Midwest appears to have lagged in comparison to other regions, peaking reserves at Q1 ’21 as opposed to Q3 ’20. Furthermore, the region as a whole has been slower to unwind reserves over 2021 until early 2022.
Finally, we look to coverage ratios for Residential Real Estate, Commercial Real Estate, Commercial, and Other Consumer segments as published to call report schedule RI-C for Q1 ’22 and Q2 ‘22. The tables above demonstrate coverage ratios and the subsequent tables outline loan composition. Consistent with other parts of our study, we’ve further segmented this data by geographic region, as well as asset size.
 
Given the high levels of uncertainty around unemployment rate and inflationary pressures, it’s sensical we’re seeing increases in coverage ratios on Residential Real Estate and Other Consumer across many institutions. Interestingly on the Commercial Real Estate and Commercial portfolios, specifically with larger institutions, we are seeing decreases in coverage ratios. It should be noted we have seen larger loan growth in these categories over the respective periods. It could be likely that “COVID” related reserves continue to unwind as Commercial and Commercial Real Estate loan types were some of the most heavily impacted by lockdowns, etc. during the pandemic. However, we continue to hear discussions of longer-term concerns around the future of Office related Commercial Real Estate loans. Therefore, as we move into future periods we may see these trends ease or even reverse.

Paycheck Protection Program Loans

New to this study, we’ve included an analysis of the impact Paycheck Protection Program (“PPP”) loans have had on reserves. As PPP loans are fully guaranteed and typically do not carry any additional reserve, including these loans in the dataset’s total balance understates the average coverage ratio. We sourced PPP loan balances for the banks who have adopted CECL from published Call Report data (Schedule RC-C, Part 1) covering Q2 ‘20 through Q2 ‘22.

To quantify the impact, we subtracted the disclosed PPP balances from the total loan balance for each institution to recalculate the quarterly ACL coverage ratios. We then compared the adjusted ACL coverage ratios to the reported data to highlight the difference in allowance. We also gathered data on the total percentage of PPP loans by region on a quarterly basis.

As expected, the period with the highest percentage of PPP loans occurred in Q2 ‘20 with an average of 6.53%. We noted a relatively high volume of PPP loans through Q1 ’21 but a steady decline in balance thereafter. Furthermore, institutions in the Southern region had the highest population of PPP loans with an average of 8% of total portfolio composition at the highest point.

Based on our analysis, the adjusted ACL coverage ratios on average were 0.04% (4bps) higher than the reported coverage ratios from Q2 ‘20 to Q2 ‘22. The period with the largest difference between the reported and adjusted reserves occurred in Q1 ‘21, where the adjusted ratios were 0.09% (9bps) higher on average than the disclosed allowance.

Since these loans have largely paid down since the program’s inception, the impact as of Q2 ’22 and going forward is relatively immaterial. That said, it’s important to keep in mind some of the other unique intervention that’s taken place as part of the COVID-19 pandemic (PPP loans, stimulus checks, loan deferments, etc.), In fact, in addition to the published allowance ratios being artificially deflated due to PPP loans being included in total loan composition, many institutions felt that calculated reserves were too aggressive due to the high levels of government intervention and consequently adjusted down expected losses through qualitative factors. Intuitively, we’d infer that reserve levels would be even higher had we not experienced such high levels of government intervention. As a recession appears more and more likely in the coming year, it’s important that institutions note these intricacies, specifically when utilizing the reserve levels in 2020 as a benchmark to periods of economic distress.

 

Methodology Selections

As our population of adopters continues to grow and the dataset captures banks of all sizes and levels of complexity, the comparability of average coverage ratios becomes increasingly difficult. There is a high level of variation among CECL adopters regarding loss methodologies, implementation processes, Reasonable and Supportable forecast assumptions, etc. While Cash Flow appears to be the most recurring model, especially for larger institutions and early adopters, the same trends do not necessarily hold true for smaller institutions. In regard to the Reasonable and Supportable forecast component of ASC 326, larger institutions are more likely to receive pressure from examiners to utilize a quantitative forecasting model with inputs based on one or more macroeconomic variables (i.e., unemployment rate, GDP growth rate, CPI, etc.). This type of model lends itself well to the flexible nature and complexity of the Cash Flow methodology. That said, smaller institutions do not always have the historical data or resources in place to develop such complex mechanical forecasts. Therefore, they may elect forecasting techniques that are more qualitative in nature by electing external economic data series and developing scale framework or scorecards that aid in populating adjustments. To demonstrate this variation in approaches taken to address the reasonable and supportable forecast, please see below for commentary from two different institutions as of Q2 ’22.

 

Huntington Bancshares, 6/30/22 10-Q 

We use statistically-based models that employ assumptions about current and future economic conditions throughout the contractual life of the loan. The process of estimating expected credit losses is based on three key parameters: PD, EAD, and LGD. Beyond the reasonable and supportable period (two to three years), the economic variables revert to a historical equilibrium at a pace dependent on the state of the economy reflected within the economic scenario. Future economic conditions consider multiple macroeconomic scenarios provided to us by an independent third party and are reviewed through the appropriate committee governance channels described below. These macroeconomic scenarios contain certain variables that are influential to our modeling process, the most significant being unemployment rates and GDP. The probability weights assigned to each scenario are generally expected to be consistent from period to period and determined through our ACL process. Additionally, we consider whether to adjust the modeled estimates to address possible limitations within the models or factors not captured within the macroeconomic scenarios. Lifetime losses for most of our loans and leases are evaluated collectively based on similar risk characteristics, risk ratings, origination credit bureau scores, delinquency status, and remaining months within loan agreements, among other factors [4].

Waterstone Financial, Inc, 6/30/22 10-Q

The Company utilized the Vintage Loss Rate method in determining expected future credit losses. This technique considers losses over the full life cycle of loan pools. A vintage is a group of loans originated in the same annual time period. The loss rate method measures the amount of loan charge–offs, net of recoveries, (“loan losses”) recognized over the life of a pool by loan segment and vintage and compares those loan losses to the original loan balance of that pool as of a similar vintage.

[…] Additionally, the weighted average remaining maturity (“WARM”) method is used for the Construction and Consumer loan pools. The WARM method considers an estimate of expected credit losses over the remaining life of the financial assets and uses average annual charge-off rates to estimate the allowance for credit losses. For amortizing assets, the remaining contractual life is adjusted by the expected scheduled payments and prepayments. The average annual charge-off rate is applied to the amortization-adjusted remaining life to determine the unadjusted lifetime historical charge-off rate.

[…] The Company’s CECL estimate applies a forecast that incorporates macroeconomic trends and other environmental factors. Management utilized national, regional and local leading economic indexes, as well as management judgment, as the basis for the forecast period. The historical loss rate was utilized as the base rate, and qualitative adjustments were utilized to reflect the forecast and other relevant factors [5].

Lastly, as many institutions have elected to use third party vendors to assist with CECL implementation, methodology selection also largely depends on the institution’s CECL vendor and which methods are available to them, ease of use, availability of data, etc. All said, as we continue to refresh this dataset and look to published reserve rates, it’s important to note the potential of increasing variability amongst institutions’ calculations the closer we get to the 1/1/2023 adoption date.

 

Regulatory Capital Ratio Elections

At the onset of FASB releasing CECL guidance back in 2016, a major concern was that CECL reserves would result in an increase in the reserve. In response, the Federal Reserve System, the Federal Deposit Insurance Corporation (FDIC), and Office of the Comptroller of the Currency (OCC) adopted a joint final rule for an optional phase-in period of three years for banks to absorb the impact on regulatory capital of implementing the new Current Expected Credit Loss (CECL) standard. On March 31, 2020, the regulatory agencies issued a joint statement to clarify the interaction between the Revised Transition of the Current Expected Credit Losses Methodology for Allowances interim final rule (CECL IFR) and the Coronavirus Aid, Relief, and Economic Security Act (CARES Act) for purposes of regulatory capital requirements. Institutions now have an additional option to the previous optional 3-year capital phase-in, mentioned above, with a 5-year capital phase-in or to delay the CECL implementation [6].

Typically, we noted that the provision was not as widely used as originally expected and that of those that took the election, most did so with the additional 5-year provision noted in the COVID relief efforts of the CARES act. In order to isolate these data points, we looked to those banks included in our study that were still active (i.e., not acquired) as of 6/30/2022. Of those 175 institutions, 121 or around 70% elected not to have regulatory capital phase in relief. Of the total 54 that did elect regulatory phase in relief, only 5 or 3% did so with the original regulatory guidelines. The remaining 49 took the relief through the CARES act provisions. From this, one could perhaps infer that if not for the impact of COVID-19, most institutions would have likely not utilized the regulatory relief efforts thus showing that the “hit” to capital that CECL produced was not significant enough to work through the regulatory reporting efforts. For 2023 adopters, we could infer that there would be a limited number of companies that would adopt for the regulatory capital relief. However, this may change should an estimated Economic Recession occur at the 1/1/2023 adoption date and prompt higher reserve ratios. The following charts detail our research.

Future Reporting Periods

As observed through review of the latter half of 2021 and early part of 2022, many banks are continuing to unwind reserves that had been built up in response to the COVID-19 pandemic. Some, however, are monitoring changes in the macroeconomic environment as we stray away from the improved conditions seen in 2021 and move more towards high levels of economic uncertainty, inflation, and talks of recession in 2022. Although the national unemployment rate has greatly improved over the last year, currently sitting at 3.5% as of July 2022, reserves are expected to increase in coming quarters as the chance of a recession is increasingly likely. In a monthly survey of economists produced by Bloomberg, the median probability of a recession over the next 12 months is 47.5%, a 17.5% increase from June’s outlook [7].

As we continue to close the gap on 1/1/2023, over 15,000 companies are projected to adopt CECL. As mentioned in prior installments of this series, although adoption of the standard can be an extremely daunting task, fortunately for this group, the industry is much more seasoned as it relates to CECL. As 2023 adopters continue to work towards their go live date, they can look to other institutions that while larger and perhaps more complex, can provide valuable insight into benchmarking, best practices for readiness, methodology and forecast selections, etc.

 

 

 

Additional Resources

[1] SVB Financial Group (SIVB)

[2] Citizens Financial Group, Inc. (CFG)

[3] Northwest Bancshares, Inc. (NWBI)

[4] Huntington Bancshares Incorporated (HBAN)

[5] Waterstone Financial, Inc. (WSBF)

[6] Regulatory Capital Phase-in for CECL

[7] Odds of US Recession Within Next Year Near 50%, Survey Shows

 

 

About the Authors

Derek Hipp, CPA

Co-Founder, COO, Director of Consulting

Derek is a Co-Founder of Valuant and leads the Client Service division. Valuant assists financial institutions with various financial modeling and data analytical services including Acquisition Accounting, Valuation, and
Credit Risk Analysis (Current Expected Credit Loss or “CECL”) through delivery of a proprietary suite of Software as a Service(SaaS) solutions, ValuCast™. The Client Service team assists customers with onboarding, training, and execution of these products and services. Derek is also a member of the Valuant leadership group focusing on team development, business development, product delivery, software feature development and exploration, and other strategic initiatives.

 

 

 

Madison Kalmanowicz

Senior Consultant

Madison has over four years of accounting and consulting experience with financial institutions ranging in asset size from $300 Million to $19 Billion. Her primary responsibilities include Day 1 valuation and due diligence services for whole bank acquisitions and branch purchases, ALLL and CECL implementation, as well as ongoing allowance advisory consulting.
Madison earned her Bachelor of Science in Finance and Entrepreneurial Management from the University of South Carolina, where she was also a member of the Capstone Scholars program.

 

 

 

 

Macy Foster

Consultant

Macy has over three years of professional experience, primarily working with financial institutions ranging in size from $1 Billion to $15 Billion in total assets. Prior to joining Valuant she worked at a “Big Four” accounting firm in Risk Advisory focusing on control testing, risk assessments, and compliance. At Valuant, Macy assists clients with Day 1 and Day 2 Accounting under ASC 310-30 an ASC 310-20, the Allowance for Loan and Lease Losses, and CECL implementation projects. Macy is a Certified Public Accountant.

She earned her Master of Accountancy from the University of Virginia and her Bachelor of Science in Accounting and Business Administration from Washington and Lee University

Download Resource

Add this resource to your reading list, or save it for future reference.

More articles you may be interested in

Ready to get started.

Let us share more about how our analytical tools and software have helped other management teams maximize their profitability…and how we can do the same for you.

Thank you!

Please help us communicate better.

As a subscriber, you’ll enjoy exclusive offers and product updates directly from Valuant. Providing additional information will help us deliver the updates you care about most. We value your privacy and promise never to share you personal information or email address.