On Jan. 1, 2020, the Financial Accounting Standards Board (FASB) put in place a transformational change to reserve accounting in the U.S. for larger public business entities that are SEC filers. (The smaller reporting companies have been granted an extension until January 2023.) The FASB replaced the incurred loss methodology under U.S. GAAP with a forward-looking current expected credit loss (CECL) methodology.

What this means is that under CECL companies will need to set aside allowances at the time they originate loans rather than once the loans are impaired. As such, CECL reserves should take into account the lifetime of expected credit losses. The FASB expects this practice will prompt bank managements to, in theory, better prepare for changing phases of an economic cycle and help avoid the undue volatility of the rearview-mirror approach of the previous methodology. (For more information, see "How An Accounting Change For U.S. Banks Could Affect Reserves And Ratings," published June 1, 2018, and "Delving Deeper Into The New Loss-Accounting Rules And Their Effect On U.S. Financial Institutions," published March 20, 2019).

As banks undergo this transition, we don't expect to take any rating actions based solely on CECL implementation; however, the change will reduce banks' regulatory capital ratios as well as S&P Global Ratings' risk adjusted capital (RAC) ratios, while increasing most banks' loan reserve allowances. We will monitor the combined impact of RAC plus reserve/loan ratios, in order to assess each individual bank's ability to absorb credit losses.

As of the start of 2020, the majority of the banks we rate have estimated the likely increase of their loan reserves due to CECL implementation. We have used these estimates to gauge how much banks' RAC ratios will decline--and concomitantly how much their loan reserves will increase. Overall, we estimate the median RAC ratios of the banks we rate will likely decline roughly 17 basis points, and loan reserves will likely increase roughly 35%. But there are large outlier banks to the median, with the greatest RAC declines concentrated among banks with a significant percentage of consumer loans to total loans.

The Amount Of Increase In Bank Reserves Due To CECL Varies Widely

CECL implementation will affect the loan loss reserves of banks we rate differently (see chart 1). The reasons for the wide variance are as follows:

• The composition of a bank's loan portfolio: Generally, the more consumer oriented, the higher the expected reserve.
• The credit quality of a bank's loan portfolio: Lower credit quality generally means a higher expected reserve increase.
• The economic outlook: Although the outlook for the economy at this stage in the cycle is likely roughly uniform across the banks we rate, if a bank were to have a more pessimistic outlook for the economy, it would have a higher expected reserve outcome, all else equal.
• Reserve modelling: The mechanics of each banks' modelling could differ. One important component of modelling that could account for differing results is the timing of when a bank believes the economy will return to more normal conditions. The quicker a bank expects the return, the lower the amount of reserve increase at the bank, all else equal.

Chart 1

Estimating The Impact On Bank RAC Ratios Once CECL Has Been Implemented

Using the reserve changes (see chart 1), we estimate the potential impact on our RAC ratios (see Appendix for the mechanics of this exercise). Regulators will allow banks to phase in the CECL impact on their regulatory capital ratios during a three-year period, but we will apply the full impact of CECL to a bank's RAC ratio on day 1. That's because our RAC projections typically extend over at least two years for investment-grade credits. Therefore, the benefits of phasing in the negative impact of CECL to RAC ratios over three years would largely be netted out in our capital projections anyway.

A few banks (see chart 2) could see their RAC ratios fall out of the range we associate with an adequate or strong capital score (7%-10%, and 10%-15%, respectively). (Note: The chart reflects the immediate impact to RAC ratios as of June 30, 2019; it does not reflect the two-year forecasted RAC ratio, which could bring the RAC ratio back into range for its current capital score.) This in turn could result in a lower capital score for this bank. However, a lower capital score will not automatically cause us to downgrade a bank, especially if we expect the combination of the RAC ratio and reserve ratio to be net higher post-CECL versus pre-CECL.

That said, we could consider a rating change if a bank's business or product strategy were to change significantly as a result of CECL. For example, banks with a higher growth rate or a preponderance of consumer loans may be disadvantaged due to CECL (they will likely need to book higher reserves for each new loan), which may prompt them to pursue a new lending strategy. Also, we would look less favorably on banks that are significantly more aggressive than peers in terms of ongoing provisioning under CECL (i.e. building less reserves than peers, all else equal).

Chart 2

The Combined Effect Of Capital And Reserves After CECL Implementation Is A Positive Credit Story

We expect that CECL implementation will likely be neutral for bank ratings because most companies will benefit from a higher combined RAC and loan reserve ratio. That's because on day 1 of CECL implementation, banks will benefit from the full amount of an increase in reserves, but retained earnings will decline by less than this amount due to the creation of a deferred tax asset (DTA; because of the timing differences of actual charge-offs versus the creation of loss reserves). Notably, when reserves increase more materially due to CECL, the combined impact on RAC and reserves will be more material.

Chart 3

Awaiting CECL Disclosure

We are uncertain at present how much banks will disclose publicly about the modelling they incorporate to derive their loan allowance under CECL. The more information banks provide, the better investors will be able to discern the reasons for differentials in loan loss reserves. Since bank management teams will need to make educated guesses about expected losses under CECL, one aspect we believe will gain market attention is the way their reserving allocations vary in response to changes in economic conditions. Greater public disclosure would help assess the relative preparedness of banks in response to changes in the economic cycle. From a ratings perspective, banks that take a more cautious view of reserves will, in our view, be better positioned for stability.

Appendix

Mechanics of RAC ratio changes

The day 1 impact of CECL will not move through a bank's income statement. For example, if a bank needs to add 100 to reserves with a 20% tax rate, the accounting will be treated as follows: reserves, a liability will increase by 100; retained earnings, will decline by 80 (100 less the tax rate); and a DTA--an asset--will be created of 20.

Thereafter, CECL will affect the income statement. Thus, for any new loans originated, the full expected loss will flow through the income statement via the creation of higher provisions. As a result, some banks may be hesitant to grow loans as quickly as they had before, given that the immediate impact of such growth would be a negative to income (but not cash flow). In addition, some banks may opt to concentrate on those loans that are more benign from a CECL standpoint (i.e. commercial versus consumer).

To estimate the change in RAC ratios from day 1 implementation of CECL, we reduced retained earnings by the increase to reserves, less the creation of a DTA (derived by multiplying the increase in reserve by a banks' past four-quarter average tax rate); we risk weighted the newly created DTA by 350%, according to our methodology. If a bank's DTA were to increase by more than 10% of adjusted common equity (or ACE), we deduct the portion above 10% from the numerator of our RAC calculation (i.e. total adjusted capital). In the current exercise, we did not apply this adjustment because it was not material based on the ranges provided by the companies. To derive our estimated RAC ratios, we forecast earnings and capital payout two years ahead. (Note: Chart 2 only reflects the day-1 impact based on a bank's June 30, 2019, RAC ratio and does not reflect the two-year forecast.) Our forecast will include the full extent of provisions on new loans generated and only include provisions for legacy loans if we believe economic conditions have changed. For all intents and purposes, our provision estimates will at least ensure that the reserve-to-loans ratio will match the day 1 impact of CECL and will move higher if we believe economic conditions will deteriorate.

Related Research

• Delving Deeper Into The New Loss-Accounting Rules And Their Effect On U.S. Financial Institutions, March 20, 2019
• How An Accounting Change For U.S. Banks Could Affect Reserves And Ratings, June 1, 2018

Jason He contributed research to this report.