(New York, NY): As the time for Current Expected Credit Loss (CECL) implementation draws closer, most credit card issuers are working feverishly to develop workable models by the relevant deadline – 2019 for SEC filers, 2020 for non-SEC filers. CECL is intended to better capture the inherent credit loss exposure within a loan by measuring the total expected credit loss over the entire life of the loan, as opposed to other time frames, such as the next 12 months.
While the CECL standard applies to lending broadly and will not be overly burdensome for many types of loans, the new standard presents an unusual challenge for credit card issuers. The heart of the problem is the calculation of the “life” of a credit card loan. Since credit card accounts are revolving facilities, determining the life of a credit card receivable requires adoption of certain modeling positions. While the industry has not reached firm conclusions on modeling conventions for CECL, issuers have been discussing the implementation’s challenges at Auriemma’s Card Accounting Roundtable, where CECL has been a continual agenda item for the last two years.
Some of the major considerations currently under review by issuers include:
Allocation of payments. While the CARD Act stipulates certain rules for the allocation of payments, issuers are looking at other payment allocation methods for CECL. Under the CARD Act, principal payments are directed at the highest outstanding interest rate balances first (a credit card loan may have certain balances at higher rates than other balances – e.g., a 0% teaser rate for a purchase, but a 16% interest rate on other outstanding amounts). While this statutory principal allocation is intended to give the most benefit to the consumer, it can needlessly complicate the sizing of the credit loss reserve. An alternative payment allocation method called FIFO (First In, First Out) would allocate principal to the oldest balance first.
Portfolio segmentation. Any credit card portfolio is comprised of thousands (in some cases, millions) of individual accounts. Some of these cardholders will always pay the minimum amount due, some will pay the entire outstanding balance, and some will pay an amount between the minimum and the entire balance. (Using the average payment amount in a portfolio-wide calculation may give a distorted result: a portfolio with 50% minimum payers and 50% transactors will likely have more risk than a portfolio where everyone pays 50%.) While the correlation between payment amount and credit risk is not perfect, there is the general understanding that minimum payers are likely to produce a higher rate of loss than cardholders making more than a minimum payment. Most issuers, therefore, will want to calculate a loss exposure on multiple segments. Segmenting into the three categories mentioned (minimum, full, partial) would seem to be the most basic level of analysis and some card issuers will stratify their accounts into numerous payment cohorts (deciles). Beyond payment amount segmentation, issuers are also considering segmenting by other characteristics: months on book (longevity of customer), credit score, origination channel, etc. The CECL for the entire portfolio would be the sum of each CECL value for each segment. Further complexity is created by combining variables – for example, payment amount and FICO score.
The current lack of industry standards or modeling conventions has created great uncertainty among card issuers. This lack of consensus makes issuer comparisons fraught with complexity. While the original impetus behind CECL may have been a desire to create additional loan loss reserves, the lack of standards may result in widely divergent loss assessments for portfolios of comparable credit quality.
In addition to CECL modeling challenges, there are other concerns for issuers, such as governance and oversight. Should responsibility for CECL rest with credit risk, accounting, both or elsewhere? Generally, capital issues are managed at the enterprise level for large, multi-line banks. For these banks, CECL governance may be distant from the individual lines of business, increasing the likelihood that the modeling does not reflect the idiosyncrasies of the credit card business.
CECL will also have an impact on credit card portfolio M&A. In many instances, portfolio purchasers will have limited historical loss information for an acquired portfolio. Since the CECL model is based on historical loss experience, an acquirer will need to look for similar quality assets to derive a loss proxy. This additional ambiguity may result in a wider range of market prices for a given credit card portfolio.
Since increases in the loan loss allowance reduce operating income, the CECL adoption may have the unintended consequence of reducing lending capacity. Moreover, for an issuer with a growing credit card business, the CECL adoption is not a “step change” function but will be continuous.
Finally, issuers are preparing for a period of “parallel testing” where the loan loss reserve is calculated under CECL and under the previous methodology. But this begs the question: why a parallel run? It suggests that comparing the outcome of the CECL levels to levels under the prior methodology may lead to further mischief. If a card issuer calculates a lower level of loan loss allowance under CECL, will regulators require the pre-CECL methodology? A loan loss “ratchet” might be in the offing, just as the Collins Amendment became a “floor” under the new Basel III regulatory capital rules.
These are just some of the considerations that credit card issuers are weighing in the CECL implementation planning. Auriemma is involved in these matters in multiple ways including, notably, its Card Accounting Roundtable, writing comment letters to FASB, as well as bespoke consulting work.
About Auriemma Group
For more than 30 years, Auriemma’s mission has been to empower clients with authoritative data and actionable insights. Our team comprises recognized experts in four primary areas: operational effectiveness, consumer research, co-brand partnerships, and corporate finance. Our business intelligence and advisory services give clients access to the data, expertise and tools they need to navigate an increasingly complex environment and maximize their performance. Auriemma serves the consumer financial services ecosystem from our offices in New York City and London. For more information, call John Costa at (212) 323-7000.